In a break from bureaucratic history, I’m pleased to announce that one of my projects at the Foundation for American Innovation just launched: the American Governance policy team’s website Plain Job Titles Please.

Federal agencies were ordered to stop using unclear job titles. The Office of Personnel Management singled out the title IT Specialist as especially bad. To track compliance, this counter goes up every day federal agencies avoid posting a job listing for an “IT Specialist.” The counter resets to zero when they use the forbidden title. A higher number is better.

OPM’s mandate about adopting plain, industry-standard job titles is important for recruiting and retaining talented civil servants. But announcing this initiative is the easy part. What really matters getting agencies to actually change their HR practices and stop trying to hire “IT Specialists.” Unfortunately, so far in 2026 the federal government has not yet gone a single business day without trying to hire one. That’s why we built this counter: to help the government follow through on a worthy initiative.

Check out the site below.

https://plain-job-titles-please.com/

Many of us have been frustrated with local government. Perhaps the roads go unpaved, the streets go unplowed, or traffic tickets are handed out too aggressively. Most of us don’t do much besides be quietly unhappy, although some angrier citizens take their complaints to city hall. An elite few judges, however, have dared strike a mortal blow against local government and attempt to abolish it via judicial decision.

This post covers the surprisingly rich American tradition of “state supreme court decisions ruling that nearly all municipal government is unconstitutional.” There are two such cases that I’m aware of, one in Kansas and one in Ohio—but, as the saying goes, it’s odd that it happened twice.

The underlying problem arose from constitutional law. Both states’ constitutions required that laws about cities be written in a generally applicable way rather than on a case-by-case basis. Both states ignored this provision and wrote city charters on a case-by-case basis anyway. Judges responded by observing that the state constitutions forbade these charters and therefore ruled that nearly all city governments were unlawful.

Although the history is certainly entertaining, it showcases deep truths about constitutional law. The costs are high, as legal compliance often has unacceptable or absurd consequences even when the constitutional provisions work as intended. The benefits are limited, as judges are able to stop unlawful action but are not suited to construct better alternatives. As a result, attempting to achieve major goals through constitutional law results in legal red tape while also neglecting the more fundamental task of making the government more competent. It is ultimately better to ask what laws would make government work well than ask what constitutional provisions would forbid it from working poorly.

The Chartering of Great American Cities

The U.S. Constitution directly establishes the existence of federal and state government. City government, by contrast, needs to be created by legislation at the state level. States do so by giving municipalities city charters that confer on them the legal right to elect officers, tax their residents, pass and enforce regulations, and so forth.

States can create these charters in two main ways. First, they can create a general framework for charters, either by directly writing laws that apply to all cities or by establishing the process for cities to write their own charters. Second, they can write city charters themselves on a case-by-case basis—we might imagine the Texas legislature passing a law that gives some particular new city charter to Houston.

The idea of writing city charters on a case-by-case basis sounds disreputable today. However, it was once commonplace for legislatures to write many kinds of law this way, which was referred to as special law or private law. This was in fact the majority of law that Congress passed well into the 20th century; Congress would, for instance, give citizenship to individual immigrants and pensions to individual widows.

Although special law lasted into the mid-20th century, there was a backlash against certain types of special legislation in the 1800s. There was particular backlash against special municipal legislation, whereby state legislatures wrote and altered city charters for individual cities.

State legislatures were resented by cities for three broad reasons. First, their special laws were often corruptly used to give favored contractors work or give political cronies jobs. Second, the special laws were often passed after municipal elections to gut the other party—if the opposite party won, state legislatures would frequently reassign all the important powers to the county (or vice-versa). Third, state legislatures were simply not very responsive to local needs and neglected regulations that cities considered urgent necessities, in areas from electrification to public health.

Reformers demanded that cities be free of interference from state legislators. This movement, later known as the Home Rule movement, had already won victories in the 1850s when states wrote constitutions that forbade municipal legislation via special laws.

These constitutional provisions were meant to curb a real and persistent abuse. State legislatures consistently meddled with cities for partisan purposes or personal interests, and banning private municipal legislation was intended to remove this temptation. However, the provisions raised an inevitable question: how should the legal system respond if state politicians continued passing special municipal laws anyway? This was put to the test in Kansas and Ohio.

A Tale of Two Cities’ Charters

The two cases were separated by several decades, although both arose from very similar constitutional provisions. In Kansas, the relevant case was Atchison v. Bartholow (1866); in Ohio, the case was State ex rel. Kniseley v. Jones (1902).

Kansas

Kansas adopted a constitution in 1859 that forbade special legislation. Its provision was likely borrowed from the Ohio constitution of 1851 (for which, see below). The constitution provided:

Art. XII, Sec. 5

Provision shall be made by general law, for the organization of cities, towns and villages, and their power of taxation, assessment, borrowing money, contracting debts, and loaning their credit, shall be so restricted as to prevent the abuse of such power.

The provision clearly banned the use of special legislation to charter cities or alter their powers. And yet, in what might be called a conspicuous lack of statesmanship, a legal historian noted that “several successive legislatures which assembled under the constitution paid no attention whatever to these provisions. They proceeded to enact special laws for cities without the slightest apparent hesitation.1”

Among these dubiously chartered cities was the city of Atchison, which by special legislation was given the power to conduct tax assessments on its residents to fund public infrastructure.

Theodore Bartholow was a disgruntled resident of Atchison who sued the city in 1866 to stop it from taxing him. He raised several arguments as to why he ought not be taxed, most of which were mind-numbingly technical. However, Bartholow’s most successful argument was his boldest: he claimed that the city couldn’t tax him because this power didn’t even legally exist, as it was unconstitutionally granted by special legislation.

This argument was, in fact, entirely correct. The state constitution forbade case-by-case laws about cities, and that was precisely how Atchison was granted this authority to tax. The Kansas Supreme Court saw no alternative but to rule in the plaintiff’s favor. It regretted the results, saying (perhaps melodramatically):

The court is aware of many of the disastrous consequences which must necessarily follow this decision; and if in conscience it could have done so, would gladly have avoided them. It had but a single duty to perform. It has endeavored thoroughly to understand the subject, and has acted in accordance with its convictions. If the conclusion at which it has arrived be erroneous, the regrets of the sufferers will not be keener than those of the members of this tribunal. But if the decision is correct, it is better that it be now declared, than that the blow should fall with greater effect hereafter.

Decision of the court below reversed.

All the justices concurring.

Theodore Bartholow did not have to pay his city taxes, because the city’s tax department didn’t legally exist! Not only were the legal foundations of Atchison constitutionally flawed, but most of the cities in Kansas were in the same boat. Years of tax assessments, regulations, and criminal convictions were potentially threatened due to the foolhardy legislature.

And did the Kansas legislature emerge humbled, with a newfound respect for the strict letter of the law?

Not in the slightest. Despite this chaos, the legislature happily continued passing special legislation for cities, notwithstanding the judicial ruling and the state constitution.

These later (and indisputably unconstitutional) city charters were only rarely challenged, for reasons that remain unclear. This is doubly odd because raising this point was apparently inevitably a get-out-of-jail-free card for tax delinquents (or whoever was suing the city government). Lawyers forgot about this constitutional provision as quickly as the legislators. One wonders if Kansas lawyers and Kansas legislators were reading their own state constitution.2

Ohio

In 1851, Ohio adopted a constitution with the ban on special municipal legislation that Kansas likely borrowed:

Art. XIII, Sec. 6

The general assembly shall provide for the organization of cities and incorporated villages by general laws, and restrict their power of taxation, assessment, borrowing money, contracting debts and loaning their credit, so as to prevent the abuse of such power.

In contrast to Kansas, Ohio made a quite sincere effort to write generally applicable laws about cities.

This is trickier than it might appear at first glance. The laws suitable for a large city are totally different from those suitable for a small rural village. Cities need laws about, say, public health and infrastructure, and they need a bureaucracy to administer them. By contrast, the complexity of this system would crush a village that tried to implement it.

Ohio hit upon a sensible solution and said that it would divide cities into four classes based on their population. The first class cities had a population over 20,000, the second class cities had a population between 5,000 and 20,000, et cetera. Each class of city then got its own municipal code, so that (for sake of illustration) laws about power plants were only applicable to the large cities that actually had power plants.

After this promising start for the constitutional provision, the succeeding generations of politicians promptly began looking for ways to circumvent it. They began creating subcategories of classes which were eventually divided so finely that subcategories contained only a single city. For instance, one class of city was those that had a population between 31,000 and 50,000 at the previous census, which only described Columbus. At first, the courts accepted these laws because they were facially neutral.

These laws were soon used for overt partisanship. It reached a boiling point in 1902 with a series of nakedly partisan bills. One reassigned the management of Cleveland’s parks to a county board, while another took Toledo’s police board out of local elections and placed appointments in the hands of the governor. City governments—particularly mayors—had their powers stripped and reassigned to either counties or the governor. All of these bills were written as putatively general laws for a subcategory that described only a single city. They were obviously special legislation by another name.

This resulted in a variety of lawsuits filed both by the aggrieved politicians and by the public. The Ohio Supreme Court issued three opinions in 1902 that overturned these laws on the basis that they were special legislation, ruling that the system of subcategorization was unconstitutional. This judgment overturned decades of precedent that had treated the subcategories as acceptable, and would have made most city governments unlawful in some form or another. (In fact, it would have totally eliminated the government of Cleveland.) The court stayed the enforcement of its own decision.

The governor summoned the legislature for a special session, where it wrote and passed a new municipal code that eliminated the subcategories and once again was reasonably general.3

Conclusion: Constitutionalism and Administration

State constitutional bans on special municipal law did—with great difficulty!—reduce state interference with cities, but they already pushed constitutional law to the limits of what courts can effectively enforce. Constitutional law is enforced by judges, and judges are good at stopping unlawful action but bad at designing replacement systems. This dynamic made the bans on special municipal legislation difficult to enforce and casts doubt on more ambitious forms of modern constitutionalism.

Constitutional provisions work well when saying “no” solves the problem at hand. For instance, suppose the government attempts to prosecute someone twice for the same crime. A judge can rule that this violates the Fifth Amendment and order that the case be dropped. Similarly, suppose that a regulatory agency attempts to pass regulations for which it lacks legal justification. A judge can rule that the regulations are unlawful and order the agency not to enforce them. In both cases, ordering the government not to do something addresses the issue without creating a mess for anyone else.

Constitutional provisions work poorly when the problem requires hammering out a new system. Solving such problems tends to require the cooperation of other actors in order to succeed, as shown by the cases at hand. When state legislatures passed special laws for cities, judges could rule that these city charters (or newly granted powers) were unconstitutional. However, judges saying “no” wouldn’t solve the problem at all—cities would be left without key powers to raise taxes or construct infrastructure. The only real solution was for state legislatures to write general municipal codes. In Ohio, the judicial decision successfully drove reform because the governor and legislature responded; in Kansas, the legislature’s continued indifference left judges without a good remedy. Constitutional law had reached the limits of what it could do without turning judges into administrators.

And by modern standards, bans on private municipal legislation were a fairly restrained goal! For all the trouble they caused, they were fundamentally a straightforward requirement that the government refrain from doing something.

By contrast, constitutions today often pursue complex societal goals, for example by granting affirmative rights to housing or to a clean environment. But what do judges do if the government ignores these provisions? Either they do nothing, which guts the constitutional provisions, or they attempt to enforce them and thereby draw the judiciary into the realm of programs, budgets, and staffing, for which it is unqualified. In short, “affirmative rights” constitutionalism suffers from the same problems that Kansas and Ohio faced, but applied to problems where the need for coordinated administration is vastly greater.

Because constitutional law is a blunt instrument, it is most effective when used to define governmental structure and to protect individuals against unlawful government action (such as prosecution or regulatory enforcement). In these cases, judges can enforce the rules through the simple remedy of telling the government “don’t do that.” But the more that constitutional law requires affirmatively constructing new solutions, the more likely it is to have fragile outcomes that depend on other actors’ support. Moreover, this approach distracts from negotiating and implementing a durable political solution.

At the end of the day, you cannot run a society by court rulings alone. Someone has to build things that work.

Works Cited

1. H.L. McBain, The Law and the Practice of Municipal Home Rule (Columbia University Press, 1916).

2. John Archibald Fairlie, “Municipal Crisis in Ohio,” Michigan Law Review 1, no. 5 (1903).

Today’s federal agencies face a broad loss of confidence in their claims to expertise. A century ago, federal agencies also operated amid intense populist distrust, yet were run by scientists and engineers who served across administrations. My new article examines how early agencies’ organization made expertise credible and how midcentury reforms eroded that credibility. I lay out the full history in Personnel is Policy: The Fabric of Government Organization, recently published in American Affairs.

https://americanaffairsjournal.org/2025/11/personnel-is-policy-the-fabric-of-government-organization/

Every problem in government has been faced before. The fastest way to understand how to tackle today’s challenges—or how to avoid past mistakes—is to study how others have solved or framed them.

This annotated bibliography collects that knowledge: the books and documents that have helped me go down the rabbit hole of bureaucracy, ranging from manuals written a century ago to contemporary histories.

I’ve drawn on several kinds of sources. Some are theoretical works on budgeting, personnel, and other core functions. Others are historical documents—manuals, reports, and guides actually used by agencies in the past. And some are case studies and histories that cover particular agencies and programs.

Basically, if it helps explain what government does and how, I include it in this bibliography. These are works I’ve found especially valuable rather than a comprehensive review of the literature. If the documents are in the public domain or posted online by the author, I provide links.

I’ll be updating this bibliography periodically and hope to expand it significantly over time.

General histories and overviews

These sources are general histories and books which cover cross-cutting issues (and therefore do not fit in other sections).

Leonard White – Federalists; Jeffersonians; Jacksonians; Republicans
A magisterial history in four volumes. Leonard White tells the history of the United States from the viewpoint of the federal government—what was it doing at each point and why? He explores the nitty-gritty of how tariffs were administered, how laws were enforced, and what it all meant for American society more broadly. Probably the greatest work of public administration. Unquestionably needs to be reprinted. However, the final volume is rather forgettable.

Daniel Carpenter – The Forging of Bureaucratic Autonomy (2001)
This book discusses how the government got its act together during the Progressive Era and built up its competence. A truly excellent book that manages to drill down into the specifics while also looking at the bigger picture. He looks at what agency-level reform actually looked like in practice, and grounds the agencies’ success in the relationships they were building with allies in congress and civil society. One of the first two books on public administration that I ever read, in fact.

Subjects

This section is broken down into different subjects, which might be functions, like budgeting or personnel, and program areas such as municipal reform or agriculture.

Budgeting and Accounting

These sources cover budgeting and accounting: how the government plans to spend money and records how money has been spent. They include both the appropriations process in Congress and the internal budgeting process of agencies.

Edward Fitzpatrick – Budget Making in a Democracy (1918)
One of the earliest contributions to budgeting. Intriguingly (but unconvincingly) defends legislative supremacism against executive-centric budgeting. Extremely valuable summary of the various proposals for budgetary systems. Available free on Google Books.

Charles Dawes – The First Year of the Budget of the United States (1923)
An extraordinary book that is essentially the diary of Charles Dawes. The US adopted a budgeting system for the first time with the Budget and Accounting Act of 1921. Dawes was the first budget director and wrote about what it was like to create a budget process from scratch. As an aside, Dawes might have been the only person ever to have felt that Harding was a highly competent and honest man. Available free on Google Books.

WF Willoughby – The National Budget System (1927)
An evaluation of the US’s then-new budget system. It mainly summarizes the budget process. Interesting discussions of several quite obscure topics: 1) odd budget gimmicks like permanent appropriations; 2) how the White House administered the budget in the 1920s; and 3) how appropriations for infrastructure evolved over time. Available free on Google Books.

JW Sundelson – Budgetary Methods in National and State Governments (1938)
A comparative study from the 1930s of how different countries did appropriations. Most other countries imitated the UK, except for France, whose budget process was a catastrophe. Quite comprehensive. Very difficult to find.

Lucius Wilmerding – The Spending Power (1941)
The definitive history of Congressional oversight of executive branch spending. A must-read, although his attack on GAO is less relevant today (as some, but not all, of the problems he complains about have been fixed). The 19th century history is especially valuable. Ought to be reprinted.

Elias Huzar – The Purse and the Sword (1950)
A history of military appropriations during World War II. It paints an interesting picture of how members of the military appropriations subcommittees actually viewed the impossible task of budget scrutiny during total war.

FC Mosher – Program Budgeting (1954)
An important contribution to the theory of budgeting. Mosher writes about a transitional era—previously, budgeting had focused on the inputs to be bought, such as salaries, equipment, etc. Increasingly, budgeting focused on the things to be produced, that is, the specific agency programs. He writes about how dry budgetary reforms changed agency incentives, and was an unheeded voice of caution. Ought to be reprinted.

Arthur Smithies – The Budgetary Process in the United States (1955)
A lengthy discussion of how budgeting and appropriations worked in the 1950s. Not key reading, but contains some important information, especially regarding the process for infrastructure appropriations.

Burkhead – Government Budgeting (1956)
The best book ever written on budgeting. He synthesized every viewpoint about budgeting: everything from the old fashioned approach that focused on typewriters and travel expenses, all the way to Keynesian stabilization policy. Particularly strong on discussing how different approaches to budgeting affect program implementation. Desperately needs to be reprinted.

Fenno – The Spending Power (1966)
The definitive explanation of what regular order for appropriations looked like back in the “good old days,” or roughly the 1940s–60s. He interviewed politicians and he documents (across 700 pages!) the entire appropriations process. It is especially strong on the incentives and motives of key players. Ought to be reprinted.

AE Buck – Public Budgeting (1929)
An excellent book, albeit very old-school. This comprehensive book focuses on budgeting in states and cities, and covers essentially everything there was to know in 1929. He even discusses the pros and cons of different sizes of paper for printing budgets, and offers advice on proofreading! Freely available on Internet Archive and on Hathitrust. (The former is more easily available, the latter is a higher quality scan.)

Society for the Advancement of Management – Federal Fiscal Series 1–7 (1945)
A series of seven case studies on how federal agencies budgeted internally for projects (i.e., not covering the congressional side of things). It discusses agencies ranging from the Coast Guard to the TVA. If you read and manage to understand this bureaucratic sorcery, you will probably know more about budgeting than some actual budget officers. Available free on Hathitrust.

Taxes and fiscal systems

These sources cover the way that revenue is raised and spent. They focus on the mechanics of how this is done. (By contrast, budgeting is the conceptual plan for what money should be spent on.)

Robert Love – Federal Financing (1931)
The history of the federal government’s process for issuing and selling debt. Surprisingly interesting—it was wildly politicized and rather incompetent well into the mid 1900s. The author has an interesting table of every loan incurred or debt issued by the federal government, from independence up through the date of publication. Some enterprising grad student should digitize his dataset and extend it.

Daniel Selko – The Federal Financial System (1940)
A comprehensive look at the government’s fiscal system circa the 1930s: how it collected taxes, made payments, and so forth. The factual content is extremely good. However, his reform proposals are uniformly bad and should not be read uncritically.

Personnel

These sources cover government HR practices: how employees are hired, trained, and fired. It includes such topics as the intellectual history of civil service, different approaches to HR, and case studies of agency recruitment.

Paul van Riper – History of the United States Civil Service (1958)
This book is distinctly of the ‘old school,’ digging deep into the EOs and public laws in order to tell the story of how the civil service arose and was institutionalized. He combats many myths with hard data. For example, the growth of the civil service was much slower and with many more setbacks than most people would imagine. Probably the best book about the US’ civil service (but now very dated). Much of the information cannot be found anywhere else. Should be reprinted.

WT Hammack – Training and Promotion in the Federal Prison Service (1941)
The Federal Prison Service was founded in the 1930s to bring order to the US’ woefully mismanaged prisons. Its founding director ran into a small problem—there were no qualified prison guards to hire, and (given the depression) they had no budget whatsoever. This book tells how the agency kept a hiring pipeline rolling and invested in its employees on a shoestring budget. Among the best case studies in government history. It also made me realize just how difficult it was to run a prison in the past. Available free on Google Books.

Gladys Kammerer – Impact of War on Federal Personnel Administration 1939–1945 (1951)
A deep dive into how WWII shaped the federal government’s personnel practices. The war led the government both to invest in its managers and to take very damaging shortcuts. Both halves are quite interesting, the documentation of successful investment and the case study of how bureaucracies respond to crises. Available free on Hathitrust.

RS Avery – Experiment in Management: Personnel Decentralization in the Tennessee Valley Authority (1954); Harry Case – Personnel Policy in a Public Agency: the TVA Experience (1955)
Two books about the TVA’s personnel practices from its golden age of world-leading engineering. Avery’s book focuses on TVA’s relationship with its unions. It was a strikingly unusual setup, probably the most European approach I have heard of in America. Case’s book covers the TVA more broadly. Both offer lessons in disciplined recruitment of technical talent. (But even by the time these books were written the TVA had baked in dysfunctions that later plagued it—see Hargrove’s book below.)

Administrative law

These sources cover the process by which government agencies can regulate individuals and corporations, and how agencies can take action against those who break laws and regulations. It also covers how these actions can be challenged, both in internal agency disputes and via judicial proceedings. The older sources on administrative law are much better than more recent work.

George Brown – Judicial Review in Customs Taxation (1918)
A brief and readable article that is one of the best works on administrative law. It advocates for more robust judicial review of agency action, firmly defending continental skepticism of state power against traditional anglo-American doctrines of sovereign immunity. Page 9 has perhaps the most amusing claim about administrative law ever made. Available free on Google Books.

Ernst Freund – Administrative Powers over Persons and Property (1928)
This work is 600 pages of comparative administrative law and is accordingly not so much a book as a tome. But what a tome it is. He starts from first principles by classifying the forms of administrative action in order to develop administrative law doctrinally He writes as a right-liberal whose heart certainly lies with strong political oversight of agencies, legislative supremacism asserted via narrowly drafted laws, aggressive judicial review, and minimal restraint on the private economy.

I think every topic of administrative law is discussed in here at some point, and he anticipates nearly every interesting later debate. Unfortunately, however, Freund was an immigrant from Germany and his writing exemplifies ponderous German academic writing. The historical sections are insightful but not nearly as good as the theoretical sections. Available free on Internet Archive.

FF Blachly & ME Oatman – Administrative Legislation and Adjudication (1934); Federal Regulatory Action and Control (1940)
If Freund anticipates nearly every interesting debate, Blachly and Oatman anticipate the rest. Blachly and Oatman were liberal New Dealers who favored agency independence and sweeping regulation; they ultimately broke with FDR over his support for placing independent agencies under greater political direction.

The two attempt to classify and categorize the different forms of agency action as they existed in the 1930s. They ground their taxonomy in the purpose of what the agency is attempting to accomplish. Like Freund, their taxonomy is comprehensive and illuminating. Also like Freund, it is so complex as to be barely comprehensible for first-time readers. Unlike Freund, their discussion of opposing viewpoints is laughably unfair. Still, a wealth of material that has scarcely been tapped.

Administrative Legislation and Adjudication: available free on Hathitrust
Federal Regulatory Action and Control: available free on Hathitrust

Robert Cushman The Independent Regulatory Commissions (1940)
Cushman was FDR’s advisor on regulatory reform; this extraordinarily readable and informative book began as a short study that he wrote for an FDR admin commission. He traces out the history of what agency independence meant to agencies and Congress, and how this understanding evolved over time. He argues that independence was never a meaningful concept. He also discusses how to reform agency adjudication, an issue still in the news today.

There’s a great deal of amusing score settling with his opponents from the 1930s—history is written by the losers. The history of the various agencies is outstanding. However, there is a long and boring section on English administrative law that can be skipped. Ought to be reprinted.

Ashley Sellers – Administrative Procedure, a Suggested Classification (1940); Administrative Procedure and Practice in the Department of Agriculture (1940)
Sellers was the counsel for the Department of Agriculture and wrote several studies about how USDA conducted rulemaking and adjudication. These studies were some of the studies written to inform the debate over the APA. Sellers’ studies are better the others because they were authored by a single person (whereas most other APA material was written by committee).

Sellers reviews the various regulatory laws that USDA enforced and writes about the nitty-gritty of how they were administered. He wrote a general summary—Administrative Procedure, a Suggested Classification—and also more detailed studies of individual laws.

Administrative Procedure, a Suggested Classification: available free on WUSTL Law Review.
Administrative Procedure and Practice in USDA: part 1, part 2, part 3 are freely available on Google books. (This list is not quite comprehensive. I have not located all of the individual studies in Sellers’ series.)

Contracting and Acquisition

These sources cover how the government purchases goods. This can range from routine IT purchases to infrastructure and major defense acquisitions.

James Nagle – A History of Government Contracting (2012)
Exactly what it says on the tin. The history of contracting tells a surprising amount about the history of government generally. The book shows that we’ve already had several cycles of insourcing and contracting as government competence has waxed and waned. Similarly, contracting has often been a broadly political issue, especially during wartime. A quite good history, although I wish he were a bit more opinionated. Available free online.

Steven Kelman – Procurement and Public Management: The Fear of Discretion and the Quality of Government Performance (1990)
An early and influential book on what we now call state capacity. The author Steve Kelman argues that the fear of collusion with contractors makes agencies so dysfunctional that the policies ironically empower contractors. Somewhat outdated, but for a good reason: Kelman later became Bill Clinton’s procurement man and fixed several of the problems he identified. Remains a great statement of the principles underlying procurement reform.

Eric Lofgren – A History of Thought in Defense Acquisition (2017)
A brilliant and capacious book on defense acquisition, an area that is a beast even by the standards of government acquisition. Lofgren ties acquisition to the broader questions of agency design and institutional culture, particularly by attacking the mechanistic ideas of mid 20th century corporate thinking. Much of his discussion is the current state of the art. Available free online.

Michael Garland – A Brief History of IT Acquisition Reform (2015)
There are many proposals for making IT acquisition more nimble, but where did the ponderous status quo come from? This brief history begins with a discussion of the prereformed system: in the 1960s, the Brooks Act imposed a centralized and bottlenecked approach that later reformers have reacted against. He then covers the repeated attempts to reform IT acquisition through undoing this centralization. Today, we swing back and forth between “centralize IT acquisition to save money” and “decentralize IT acquisition to move faster.” Garland shows that this dynamic has always existed.

Available for free online on Garland’s website.

Gerald White – Billions for Defense: Government Finance by the Defense Plant Corporation During World War II (1980)
A study of a niche aspect of World War II: the way that new plants for defense production were financed by the government. The government unsuccessfully attempted to stimulate investment with tax breaks and the like until turning to a more robust approach: directly loaning funds for building plants. This was carried out by a government sponsored enterprise, the Defense Plant Corporation. Worthwhile reading for the debate on industrial policy today.

Industrial Mobilization for War, History of the War Production Board and Predecessor Agencies (1947); The Army and Economic Mobilization (1957); Study of Experience in Industrial Mobilization in World War II (1945)
These are the main documents about the US’ industrial policy during World War II, namely retrospective studies written soon after the event. All of these are massive and not for the faint of heart. However, the amount of information on industrial mobilization contained in them is unparalleled.

Industrial Mobilization for War: Available free on Hathitrust
The Army and Economic Mobilization: Available free on GovInfo (and elsewhere)
A Study of Experience in Industrial Mobilization in World War II: vol 1 and vol 2 freely available on Hathitrust

Management

These sources cover ways in which agencies attempted to become better managed, particularly through training their managers in process improvement.

US Department of Agriculture – A Guide to Better Management (1944)
A series of case studies on improving government process, compiled by the US Department of Agriculture during World War II in order to train its employees. The case studies include everything from major reorganizations, to improving internal documentation, to streamlining payments to contractors. The document is especially valuable in that it frequently includes the actual memos that managers and employees wrote when they identified and fixed problems—rarely is bureaucracy made so human. Not for everyone, but it is the best at what it does. Available free on Google Books.

US Department of the Army – Manpower Utilization: Report on Work Simplification (1944)
A document on work simplification, a method for improving bureaucratic process, written during World War II to train Army officials. The Army pioneered this approach before it was later rolled out to the entire federal government by the Bureau of the Budget (now OMB). This document shows the origins of the program and includes several techniques not covered in other documents. In particular, it discusses ways of improving the layout of factories and storage facilities. Available free on Google Books.

US Bureau of the Budget – An Agency Management Program (1944); Public Administration Service – Work simplification as exemplified by the work simplification program of the U. S. Bureau of the Budget (1949)
During World War II, the civilian government was nearly crippled by the exodus of trained employees to the national defense program. In response, the Bureau of the Budget (now OMB) launched a training program for managers to help them improve their agency’s procedures, thereby dealing with the lack of experienced personnel. Both of these documents are the actual documents used during the war.

The document An Agency Management Program was a self-assessment checklist circulated around the government. It’s brief, entertaining reading. How would government agencies stack up today? Perhaps not as well as could be desired.

Meanwhile, the work simplification document is the meat of it—a compilation of training material written and used during World War II. It also has charming illustrations. Not only a great source of inspiration for process improvement, but a great piece of Americana.

An Agency Management Program: Available free on Google Books
Work simplification: Available free on Hathitrust

State and local government

These sources cover subnational government, i.e., state, county, and city government. In particular, these sources discuss the ways in which local government is organized and the means whereby its competence is built or destroyed.

Peverill Squire – The Evolution of American Legislatures (2012)
Squire’s excellent book discusses the history of American legislatures, particularly state legislatures and their professionalization over time. Legislatures that began as part-time meetings of prosperous farmers have become sophisticated organizations with institutionalized support.

Jon Teaford – The Rise of the States (2002)
This great book covers the history of state government and how its responsibilities and competence changed over time. Contrary to popular impression, until the mid-1900s states were unimportant relative to the federal government and to cities. Teaford traces out how states tackled new responsibilities over time and gradually professionalized.

If you read Squire and Teaford, you’ll have an outstanding knowledge of state government.

New York Bureau of Municipal Research – The Constitution and Government of the State of New York (1915)
An extraordinarily influential study written for the New York constitutional convention of 1915. This was the main call to arms for the conservative reformers, focusing on the need to strengthen the chain of command of the executive branch. The study begins with a statement of the principles of executive/legislative relations and the principles of good executive management, then proceeds to offer suggestions for reforming every agency and program of the New York state government. This slept-on document set the agenda for reformers for several decades. Available free on Hathitrust.

Chester Maxey – Readings in Municipal Government (1924)
Local government reform was one of the key issues during the early 20th century. As it was a heated debate, there was much more creative thinking than today. This volume collects key early readings about municipal reform—different methods of organizing cities (and the case for each), the debate over municipal civil service, and much more. A fair amount of it is outdated, but many of the problems it highlights are evergreen issues. Available free on Google Books.

History of agencies

These sources are histories of specific agencies, particularly histories that explore how institutional factors lead to success or failure.

Erwin Hargrove – Prisoners of Myth: The Leadership of the Tennessee Valley Authority (1994)
This outstanding history of the TVA asks an intriguing question: how did an agency go from being world leading to a mismanaged catastrophe over several decades? Hargrove posits that the freewheeling culture of the 1930s and 40s laid the groundwork for later failure as the TVA tackled new problems it was ill suited for. The book implicitly makes the case that rigorous management is a crucial ingredient of bureaucratic success—vision and flexibility minus oversight led to the agency becoming woefully dysfunctional.

What else?

If you know of other essential works, or if you’ve used any of these in your own work, I’d love to hear about it in the comments.

Editor’s note: State Capacitance is pleased to present How Process Charts Kept the Lights On, by a guest contributor who goes by the name F. Ichiro Gifford and maintains the blog Energy Crystals. He recounts how he applied Work Simplification to his municipal utility. The piece offers a model of how Work Simplification might be applied in the digital era.

How Process Charts Kept the Lights On

On 23–24 June 2025, New England saw a sudden heat wave that forced our local grid operator to take emergency countermeasures to prevent a brownout. My job as a resource analyst at a municipal electric utility is to keep our costs down through engineering-informed financial analysis. High-load days like this spike our internal costs, so if I didn’t handle this, our customer’s electric increases would increase. So I ran a process to reduce our total load during the heatwave.

It was a mess. Fingers were pointed, wires were crossed, and decisions were made at the last minute.

I suspected I had screwed up. Certainly, I thought, this could be done better. I looked for instructions, but I only found disorganized scraps. I asked who ran this program before me—that person had left the company. I was dealing with managerial debt: outdated and poorly documented processes that slow down an organization, akin to technical debt. Procedural knowledge was neither recorded nor optimized—instead, it lived in people’s heads, including employees who left long ago.

I wanted to fix this, but at first, I didn’t know how. Then I remembered State Capacitance’s discussion of Work Simplification: a program started by the Bureau of the Budget in 1942 that offered management training workshops, complete with a handy-dandy pamphlet and a set of posters in an art style we now recognize as “the Fallout look.” These workshops, pamphlets, and slick-looking posters trained bureaucrats—people like me—to face managerial debt head-on. Not only did it train them in specific methods for improving processes, but it offered a better vision of operational excellence.

These pamphlets are no joke. I used one of these tools—the Process Chart—to improve this peak response process at my day job.

It worked.

The Process Chart

The Process Chart is a standardized form used to break out business processes into individual steps with type classifications. A manager would write a Process Chart, step-by-step, for any business process in his office—large or small, important or not. Each step would get classified into one of four buckets. These four buckets were written for an era in which staffers literally moved stacks of paper between desks, so I’ll give you the updated set that I developed:

Large Circle (O): Operation. Creating, changing, adding to something.

Small Circle (◦): Message. Emailing, messaging on Teams, pinging on Slack. Also make a note of who gets the message.

Triangle (△): Storage. Saving a file. Also make a note of the file path where files are stored.

Square (▢): Inspection. Checking and verifying, but not (typically) changing.

The manager lists each step of the process with the following information:

• The step’s category: Operation, message, storage, or inspection.

• What the step entails. The more specific the better.

• Who performs the step. Name backup staff where applicable.

Simple Enough, Right?

Of course not.

Let’s look at the “before” chart, adapted slightly for public consumption. We can call this process the “Peak Shredder” program—I’ve learned from customer programs that having a funny name is half the battle.

This process currently involves: me, the Resource Planner (R. Planner); the Integrated Resources Director (I.R. Direct.); the Customer Service Representative (C.S. Rep.); the Key Accounts Manager (K.A. Mgr.); the Customer Service Director (C.S. Dir.); the Control Room Operator (C.R. Oper.); and the General Manager (Gen. Mgr.).

Wait, Is This The Process?

The Work Simplification pamphlet recommends that, after creating an initial Process Chart, you should stare it down and ask who-what-where-when-why. I’ll just quote the pamphlet here:1

Remember you are looking for ways to eliminate, combine, rearrange, and simplify steps. For that reason, question each detail of the process. Ask yourself what purpose each one serves. So ask yourself these questions:

• What is done? What are the steps? Do I have them all? What does each step do? What are the surrounding facts?

• Why is this step necessary? Can as good a result be obtained without it? Is it an absolute must? Don’t be misled by an excuse when you are looking for a reason.

• Where should this step be done? Is it done in the right sequence? Can I combine or simplify by moving this step ahead or back?

• Who should do the job? Is the right person handling it? Or is it more logical to give it to someone else?

• How is the job being done? Can it be done better with different equipment or different layout? Can I make this job easier for everyone concerned?

With that framework in mind, let’s identify some problems in the Peak Shredder process.

Some of This Doesn’t Need Spot Checks

The Process Chart makes this issue immediately apparent. Count ‘em: seven emails! A big contributor to that count is (frequently unnecessary) reviews by leadership. When I walked through the Process Chart in a meeting, I learned that the emails the Customer Service Director signed off on were templates. The Customer Service Representative generated them simply by changing the date and time and then running the Mailchimp. They hardly needed the Director’s attention, let alone his signature.

Similarly, the Integrated Resources Director had been needlessly approving the peak window, despite the fact that the process had already been automated. Most electric utilities have built-in tools for predicting demand peaks; either we get that tool bundled with some utility control software, or we write a tool ourselves. The Integrated Resources Director doesn’t need to look at the same dashboard I am looking at and say “…Yup.” He can simply receive the notification email like everyone else.

Some of This Can Be Automated

During the initial peak event, I learned that I actually didn’t know how our distributed energy resource management system (DERMS) worked. We slapped together a passable dispatch schedule ourselves, but as I built my Process Chart, I learned that the DERMS had tools for automating this. The automation tool could receive our peak window messages, autonomously calculate a dispatch schedule, run it without our involvement, and auto-dump an event log into a file path of our choosing. I no longer had to do this by hand. I barely even had to look at it!

This, ultimately, meant that I could take one more task off my plate. Which was for the best, considering how much of this process turned out to be my responsibility.

Some of This Can Be Standardized

During that same Process Chart meeting, I learned that the messages sent to key accounts are pretty much all the same: “Hey, a peak is coming at this time, you know the drill, call me if you have questions.” In practice, these emails and phone calls were done informally by whoever had the closest relationship with a given facilities manager. These key account peak notifications can become a form email. In fact, we can use the same form email we send to residential customers! These key accounts do, in fact, know the drill. If they really have questions, they already know whom to call.

We had inadvertently replicated a management technique from yet another 1940s government initiative: Forms Control, a method for reviewing paperwork in order to streamline processes. In our case, looking at the template emails alongside the operational instructions enabled us to consolidate our communications down to one email that we sent to everyone, both residents and key accounts.

By this same principle, we found that the after-action press release was also amenable to a template. This means the General Manager doesn’t need to write a new one each time—I can simply run a template, cut and paste a few numbers, and let the General Manager sign it.

Tying Everything Together with a Functional File

All told, the new Process Chart looks something like this:

Much better—we managed to cut out nearly half the steps and emails!

Moreover, simplifying the Peak Shredder program gave us more than just a Process Chart. It led to better documentation of each step that we improved. After the Customer Service Representative showed that some emails could become form letters, she wrote one up and sent it around for review. After I asked the Control Room Operator how to control a battery we use for the program, he graciously wrote up an instruction manual for it.

All told, the simplification effort taught us to think operationally about the Peak Shredder program, which led us to independently recreate what the Forms Control manual called a functional file. The core insight from the manual was: because a process consists of gathering, operating on, and communicating information, an operationally excellent organization ought to consider all those documents at once. The functional file is a collection of everything related to a process.

In our case, the functional file held every form, every R script, and every template letter. Just like in the 1940s, locating everything about a process in one place enables employees to review and improve it, such as by identifying duplicative forms. As the original manual puts it:

Forms are a reflection of the work methods, operating procedures, and management know-how which give rise to their use. If an agency’s forms constitute a simple, orderly plan showing clear and related purposes, there is reason to believe that personnel—knowing whatever they are doing and why—may be giving reasonably efficient service. If, on the other hand, its forms constitute an unintelligible tangle of red tape, it is pretty safe to assume that its methods and procedures—its service to the public—are in much the same shape.

Finally, I had to find a good home for this process file within a graveyard of inconsistent permissions, poorly-labeled folders, and legacy Excel files from 2005. My organization’s managerial debt showed itself again in our file management system. But that is a problem for another day.

Operational Guidance, Not Top-Down Oversight

My experience with Work Simplification and Forms Control was a chance to address managerial debt at my utility. But even more than that, it made me understand debates at the federal level about punitive and legalistic approaches to controlling the bureaucracy (such as the Paperwork Reduction Act). They’re doomed from the start, and for the same reason that “strategic guidance” never works: they’re written by people who refuse to understand operational practice.

Top-down dictums to “just do it better” never work because the people at the top can’t lead operational improvement. They don’t know what needs to be done, because they do not (and should not) wade into the muck of front-line processes. Ideally, strategic leaders (and members of Congress!) would allow operational staff—project managers, supervisors, field officers, people like me—to iron out the devil-harboring details of their Big Ideas. But instead, these leaders hand down needless impositions and wasteful spot checks, as if they could do my job better than I can.

Operational staff don’t need dictums. We need tools. We need best practices. We need training on Work Simplification, on Forms Control, on being better managers. This isn’t rocket science—these materials were prepared and disseminated in the ‘40s and ‘50s, and they worked. Government bureaucrats weren’t more talented or harder-working back then, they were managed better. And they were managed better because their managers were taught operational excellence and pushed to achieve it.

We can only fix government dysfunction with real accountability. Pushing vague modernization initiatives from on high is not accountability. Requiring permission to send a survey is not accountability. Throwing a fit when a staffer suggests a cleaner approach is not accountability.

Teaching freshly-promoted managers how to achieve operational excellence and then letting them do it—now that’s accountability!

Editor’s Postscript

The author has kindly made available the Excel template he created for Process Charting. You can view it here. It will open in Google Sheets, so download it as an Excel file (the .xlsx file extension).

Some people claim that hard times make strong men. Perhaps it is true. In any event, World War II was a hard time that made organizations stronger, where the war demanded peak efficiency from agencies such as the Social Social Security Board (now the Social Security Administration).

The war not only made the Social Security Board a strong organization, it made it a lean organization, too. The Board invented, and practiced, ideas about continuous improvement decades before businesses adopted them as part of the lean manufacturing movement. Its management was not only ahead of its time, but it was more modern than ideas pushed by later government reformers—and far more modern than federal management even today.

Subscribe now

The Board crafted this management style—adapted from factory management training—due to wartime necessities. It needed to dramatically streamline its operations due to shortages and turnover, and pursued two initiatives to do so. First, it aimed for immediate improvements with its “why survey.” Second, the Board ambitiously pursued a new mindset of constant improvement through its permanent work simplification. These initiatives not only dealt with the issue at hand, but even anticipated the management style today known as lean manufacturing.

Today, the US government again faces grave challenges, but we lack the clarity and can-do spirit of the Board’s leadership. Across the political spectrum, observers highlight problems with the federal bureaucracy without proposing concrete steps to deal with them. If they want to actually deal with the problems they identify, they should study the wartime Social Security Board, which believed that government problems could be tackled and that better procedures could be taught.

Social Security goes to war

During World War II, the Social Security Board suffered from the shortages and turmoil of the wartime economy, as did all civilian organizations. Its executive director Oscar Powell (pictured above) said:

In one large branch of the organization, last year’s turn-over was 100 percent. Like other agencies, the Board has also had to face actual or impending shortages of supplies and equipment and also the particularly severe shortages of office space and housing in Washington.

The wartime turnover put extreme stress on the Board through ruthlessly exposing inefficient procedure. In peaceful times, if organizations have inefficient procedures their employees eventually learn the unofficial shortcuts as they gain experience. But during the war, employees left as soon as they learned the ropes, making any inefficient procedures a permanent handicap. Procedures therefore had to be flawless if the Board wanted operate even tolerably well.

To hone its procedures, it undertook work simplification, a management style I have written about previously. Work simplification trained managers in iteratively simplifying procedures to improve operations; it was adapted from the training that factory managers received.

The Board’s work simplification initiative operated on two levels.

First, the agency pursued short-run victories by reviewing its major procedures top-to-bottom, which culminated in its “why?” survey. This allowed it to eliminate needless data collection, reduce procedural bloat, and transfer excess equipment to the military.

Second, it pursued long-run transformation aimed at institutionalizing this nimble mindset, thereby developing its permanent work simplification procedures. Whereas traditional work simplification made procedural improvement the job of all middle managers, the Board made procedural improvement the job of all employees whatsoever. To accomplish this, it trained all employees in the skill of simplifying procedures.

This innovation—continuous improvement by all employees—is precisely the same kaizen method that was popularized as part of lean manufacturing.

Do ask me “why”

The Board had prided itself on its focus on operational excellence, feeling that the organization’s youth gave it vigor and a sense of mission. When the war began, the agency’s leadership accordingly leapt into action: it swiftly decentralized and simplified its operations and began identifying needless equipment. The leadership discovered, for instance, 280 telephone extensions that were no longer used, and concluded that 375 file cabinets and 220 typewriters could be transferred to the war effort due to underutilization.While this was a good start, really tackling the challenge required procedural improvement. Its leadership stated:

The problem is to identify and do away with the habits and practices which have grown up accidentally, or which, having once been useful, are no longer warranted at this time.

The Social Security Board’s most sophisticated effort was from its Bureau of Old Age and Survivor’s Insurance, which solicited suggestions for improvement in its “Why” Survey. This survey reviewed more than 50 major activities of the Bureau, asking the employees—everyone from top management to new hires—which steps were needlessly complex or even wholly unnecessary.

Starting in the fall of 1942, the Bureau spent six months examining its methods. It heard more than 6,500 suggestions by more than 2,000 employees. The Board adopted proposals ranging from large, such as closing or consolidating field office, to seemingly small, such as eliminating data collection it realized it could automatically generate. But at the scale of a government agency, even minor tweaks saved serious money. By eliminating the needless data collection, the agency saved $218,000 in printing and personnel costs—about $4.5 million, today!

The Bureau of Old Age and Survivor’s Insurance saw results: as its workload increased by 6.1%, it asked for merely 1.7% greater funding for administrative overhead to handle it. However, the Social Security Board felt that it couldn’t depend on one-off initiatives. These impressive results had to be institutionalized.

Train once, simplify twice

To lock in this new mindset, the Board turned to work simplification. However, they did not use the standardized work simplification training material that I have previously written about. They had a good reason—it hadn’t yet been written! The governmentwide work simplification course was still being created by the Bureau of the Budget.

Since the Board was developing its approach from scratch, it studied the original sources that work simplification adapted, namely the training methods for factory workers called Training Within Industry (TWI).

The Board principally studied Job Method Training, which was a TWI method for simplifying tasks. Job Method Training takes some concrete operation—in factories, perhaps welding two parts together; in government, perhaps handling a certain form—and breaks it down into a list of granular steps. The employee then goes down the list and asks why each step is necessary, what its purpose is, what best way to perform it is, and so forth. Through answering these questions, employees were then able to rearrange (and sometimes eliminate) these steps, thereby eliminating complexity to achieve the greatest possible efficiency.

So far, this is precisely the same as the standard governmentwide work simplification material. However, the Social Security Board decided to adapt in a creative way—and an even better way than the (later) governmentwide training.

The Board’s director summarized its twist on Training Within Industry:

Naturally, that program has had to be adapted somewhat; the biggest change we have made has been to enlist the whole staff—top, bottom, and middle—rather than to make work simplification the job of supervisors alone.

Traditionally, Job Method Training was used to train factory managers; similarly, work simplification was ordinarily used to train government managers. The Social Security Board, by contrast, decided to train every last employee and developed a method of instruction to do so.

First, a supervisor would call a staff meeting with his reports and explain the purpose of Job Method Training, and ultimately introduce an instructor sent by the Board’s leadership. Each subordinate would pick a part of his office’s work and, using the trainer’s coaching in Job Method Training, suggest an improvement to the office’s procedure. At a subsequent meeting, each subordinate presented his suggestions to the manager and the instructor, leading to a group discussion about the suggestions’ feasibility.

The team would then actually enact every feasible suggestion and incorporate the revised procedure into its daily work. The Board found that each session resulted in many good suggestions—so good that, frequently, several individual suggestions had each saved more money than the entire training session had cost. Its director summed up:

This process may seem to require a large investment of time. It does. But the returns on our investment so far are so large and so immediate that we want to invest all we can at such a rate of interest.

This training was intended not as a one-off initiative, but as a permanent new way of working. For instance, the Board created forms for employee suggestions and judged managers by their ability to elicit them. The new mindset was intended to be:

a continuing way of operating the Board wherein every supervisor seeks the cooperation of his subordinates in finding constantly better and simpler ways of accomplishing the work of the unit.

With this new approach, the Board stood ready to face the challenges of total war.

Lean training emigrates

The history of the Social Security Board during the rest of World War II is not especially interesting, but the history of this training after WWII is very interesting.

The Board had borrowed its Job Method Training, as stated, from the broader approach of Training Within Industry. TWI died out in American industry almost immediately after the war, while it continued in government (under the name of work simplification) up through roughly the 1960s.1

However, TWI was a big hit abroad, particularly in Japan. Companies such as Toyota used TWI’s approaches (such as Job Methods Training) to adopt a nimble, iterative approach to improving manufacturing operations called the kaizen method. In this approach, companies aim to create “a culture where all employees are actively engaged in suggesting and implementing improvements to the company” through constantly streamlining manufacturing operations one step at a time—particularly by drawing upon shop floor workers’ knowledge.

These foreign manufacturers eventually ate the lunch of established American corporations. As a result, their management style of lean manufacturing (which included more than kaizen) was, by the 1980s, held up as a paradigm of exemplary management.

The Social Security Board’s practices were not merely reminiscent of kaizen, but something that can reasonably be called an independent invention of it. Both the Board and Japanese manufacturers began with the same material, namely TWI training methods. The Board adapted TWI in much the same way that Toyota did: allowing every entry-level bureaucrat to improve processes is roughly equivalent to Toyota’s famous ability for any worker to stop the assembly lines when things went awry.

The Board even spoke about its management in a way that sounds like platitudes from the 1980s, with its director saying:

the job all of us must do and keep on doing can be done only from within and by the whole organization, from bottom to top. It can be done only by enlisting the interest and brains and efforts of all the people who do the job and so know it better than anyone else. No detail is too small, and no activity too large, to be worth critical review, reappraisal, and the indicated action.

The Board shared feedback from its employees that once again would have been right at home in 1980s promotional material. The quotes included:

“I had always thought the supervisor was supposed to do all the planning of how to do the work. Now I feel it is part of all our jobs.”

“The thing that meant most to me was that the supervisor and others were interested enough in us to even ask our opinions.”

“I learned a lot about my own work that I didn't know before and I learned more about the other people's work. That made me see just how my job is important.”

“Always before I didn't know whether we were doing the job just as we should. We were sort of turned loose to do them the best way we could. Now we know just how to go about all our work and we know it is the best way—that we helped work out and that is officially approved.”

The Board’s director recounted one supervisor’s view of the process as a whole:

"It's unlike any reorganization we over had before," he said. "This one is from the bottom up. Everyone's in on it and is interested in making it work."

While Social Security was, of course, not actually a lean manufacturer (or any other kind of manufacturer), there were striking parallels to the kaizen idea of iterative improvement. The Board had used exactly the same sources that Japanese manufacturers used, had adapted the sources similarly, and even talked about management in the same way.

The Board had anticipated this major concept of lean manufacturing—right down to analogizing this method to being lean! The Social Security Board’s article on this management initiative ended by saying, “We must take off fat, for there is no fat on racehorses.”

Lean government, then and now

Today, the government’s need to operate in the face of increasing criticism and decreasing resources will require creative approaches. The Social Security Board’s playbook could still work. It can’t be imitated blindly: today’s challenges are not those of total war, and organizations and management have changed drastically since the assembly line era. But even if its methods cannot be adopted wholesale, we should still copy its approach—figuring out how to train every employee in improving his office’s work.

As there have been (and yet shall be) many harebrained schemes for revamping government training, it is worth highlighting three key points about the Board’s training: First, the training was comprehensive, as it taught every employee whatsoever to use his knowledge for procedural improvement. Second, it was a standardized training course that didn’t depend upon external consultants and therefore could scale. Third, it was institutionalized through standardized forms for suggesting procedural improvements and by requiring managers to elicit these suggestions. These three elements assured that the new culture would actually take root rather than prove a passing fad.

The main remaining questions for would-be reformers today are: 1) what approaches should civil servants should be taught, and 2) how would agencies use those skills to streamline their procedures?

For the former, the 1940s approach of Work Simplification is a worthy starting point, but there have been many new approaches since then. These include, for example, approaches for assuring that products are functional and intuitive, called user experience; and frameworks for rapid prototyping and development that (in software) are called agile development and which (in manufacturing) are associated with companies such as SpaceX.

The second question is trickier. It is impossible to say exactly what such training would look like, or precisely how it would be used—the whole point is that it needs to be carefully adapted to government, not copied mindlessly from business. But it might include, for instance, applying ideas about user experience to an agency’s own tools and software. Front-line workers at Medicare and the VA could flag buggy software that causes errors, and managers could be held accountable for addressing the problems.

Or as another example, the agile vision might suggest judging government projects on the time needed to get to a first, failed prototype of a new service or project—akin to SpaceX quickly trying prototype rockets that explode on launch. Extensive consultation or needlessly detailed contractual requirements shouldn’t just be unnecessary, they should be a failure that harms a manager’s career progression.

Finally, this history offers more than a playbook for better training. It offers a way of approaching the problems that we face today. The Social Security Board’s director claimed that the greatest obstacle facing the government was:

weariness which comes from knowing that we are in the midst of world changes about which we can do little individually while we must continue to carry on our responsibilities and must bear eventually some unknown brunt.

And this was what the Board attempted to confront—when everyone felt helpless in the face of world events, the Board attacked feelings of helplessness at work. Ownership and true responsibility was the solution to the helplessness induced by stultifying proceduralism.

Today, many people again feel that we face “world changes about which we can do little individually.” We can respond by adopt the spirit of wartime America. Today’s challenges are a chance to give every government employee newfound ownership over his bureaucracy’s performance, and for the government to operate as efficiently as lean and nimble manufacturer.

Appendix: Sources

The main source is the article Adjusting Administration to Wartime by Oscar Powell, the director of the Social Security Board whose picture adorns this article. It is in the public domain and readily available at the link above (courtesy of the Social Security Administration). The article is rather short and very much worth reading.

All other citations are given via links.

Sources:

1. Powell, Oscar M. “Adjusting Administration to War Time.” Social Security Bulletin, November 1943.

Editor’s note: State Capacitance is pleased to present The Solicitors Against the Circuit, a guest post by longtime friend Jameson Payne. Jameson is a graduate student at Hillsdale College's Van Andel School of Statesmanship. He formerly worked at the Heritage Foundation and New Civil Liberties Alliance.

The Solicitors Against the Circuit

Who regulates the regulators?

In the American administrative state, agencies are governed by byzantine rules. A portion of these rules are statutory, being written in laws such as the landmark Administrative Procedure Act (APA). But the larger portion is based on judicial interpretation of statutes: even the APA—a mere nine pages long!—leaves much to the discretion of judges. And in this struggle to be the master of the administrative state, one institution has long reigned supreme: the U.S. Court of Appeals for the District of Columbia—better known as the D.C. Circuit.

How did it manage to win out against other institutions to gain this supremacy? Was it due to the necessity of centralization, or perhaps to the convenience of proximity to the capital? These things helped, to be sure. But I’d like to propose that another factor first built up the Circuit’s influence and later tore it down—namely, institutional competence.

The D.C. Circuit’s bar—the advocates who practiced before it—was more competent than the Supreme Court’s bar during the twentieth century, and so the court garnered prestige and deference to its administrative-friendly judgments. This changed when, over the last few decades, an equally professional counterweight arose: the state Solicitors General. These competent lawyers generally opposed the administrative state, and were able to meet the D.C. Circuit blow-for-blow. Ultimately, this helped lead to the Circuit’s longstanding doctrines of administrative law being overruled.

The D.C. Circuit and Admin Law

As most readers are aware, Congress does not spell out how agencies should operate day-to-day, leaving them to work out the details. These agencies do so by making rules that affect policy, ranging from rules about how the agency will manage itself to rules about how it will enforce and adjudicate laws.

Over time, a set of laws and judicial precedents have emerged to control agencies, including limitations on their authority and procedural requirements that they must follow when they act. This body of law is called “administrative law,” and whoever makes this law wields immense influence over a range of policy issues, from immigration to insulin prices.

While there are some statutes that control administrative law, much is left to judicial determination. So which court decides what agencies can and cannot do? There are two basic options.

One, the traditional model, lets the various district and circuit courts throughout the US make their own precedent, which is binding only within their jurisdiction. Then, if courts disagree on an issue, it can undergo appellate review by the Circuit Courts and eventually the Supreme Court. If SCOTUS settles the issue, then it is binding nationwide. This has the downside of creating less consistent administrative law across the nation when different courts disagree, but has the benefit of fleshing out the law more comprehensively before appellate courts decide it.

The other model is to delegate administrative law to courts of specialized jurisdiction, much as bankruptcy cases are handled by specialized bankruptcy courts. This is usually done by specifying one court as a venue for all litigation done under a statute, or by giving a court the authority to issue decisions that preclude litigation in other courts. The expectation is that this court will become expert in its subject-matter area, and therefore more likely to make informed judgments. It also allows for a more uniform body of law, as there is no rival court to disagree with it.

Since the rise of agencies in the early twentieth century, Congress has opted for a mixed approach: some administrative issues have been delegated to one specialized court, while many others follow the traditional model of being heard by the regional courts throughout the US.

When Congress chooses to entrust administrative issues to a specialized court, it unsurprisingly favors the capital’s appellate court, the D.C. Circuit. With many statutes specifying the Circuit as the court-of-review, the D.C. Circuit holds a premier status in shaping administrative law for the whole nation.

Many of the most significant administrative-law developments in the 20th century trace their origin to the D.C. Circuit. The controversial practice of “universal remedies” was pioneered by the Circuit: it was issuing nationwide injunctions all the way back in the 60s, and universally vacating agency rulemaking was the Circuit’s “ordinary result” in a successful APA claim by the 80s. It also played a decisive role in expanding standing—who can challenge agency action—and ripeness—when agency action can be challenged. We even owe the infamous doctrine of Chevron deference to the Circuit’s hand. As Tom Merrill details in his book, what was originally an innocuous opinion applying routine canons of statutory interpretation was transformed, by the D.C. Circuit, into an overhaul of the relationship between courts and agencies.

The D.C. Circuit crafted, through these doctrines, an administrative law where the Circuit could aggressively set aside actions by democratically elected politicians, invite (mainly liberal) interest groups to sue agencies, and demand that other courts defer to the actions taken by these putatively expert agencies at the behest of interest groups.

Not everyone was happy with the D.C. Circuit’s preeminence and the results thereof.

The D.C. Bar: A Crash Course

Reagan’s Secretary of the Interior, James G. Watt, once lamented that “eastern judges (judges who are 'foreign' to many values of the West) are making decisions which control our destiny as a people and as a vital force in America's economy.” The D.C. Circuit had come to represent a bastion of elitism to its detractors; to its defenders, a desirable venue for efficient and centralized development of administrative law. But both its defenders and detractors could have agreed (although perhaps not happily agreed) on this: the court was highly competent.

Competence, however, means more than that the court had the best judges with the most experience. It also has to do with the quality of the bar—the people who practice before the court. It was an open secret among admin law gurus that litigation before the D.C. Circuit was generally of higher quality than even the Supreme Court. In the 1980s, D.C. Circuit Judge Wald confirmed the folk-wisdom: “[i]t has been suggested by some Supreme Court Justices in recent years that the quality of advocacy in our court generally is a match for-or even better than-that in the Supreme Court.”

Even such an important case as Roe v. Wade illustrates this: Texas was represented by assistant attorney general Jay Floyd, who opened his statement about abortion with a quip about the attractiveness of the (female) opposing counsel. Both parties then argued so poorly that Justice Blackmun cited it as a reason to rehear the case. Nor was Floyd’s disastrous performance atypical—Blackmun privately ranked all Supreme Court arguments; Floyd got a “B”.

This might come as a surprise; after all, the stakes of winning at SCOTUS are much higher, and so one would want to invest in the highest-quality advocates possible. But well into the late-twentieth century, the Supreme Court bar was not especially professional. A case would often be argued by the attorney who handled the case at trial and who then appealed the results—small town lawyers would end up at SCOTUS.

Hypothetically, the Supreme Court had rules that precluded such unqualified lawyers. But these rules were frequently waived via a grant of pro hac vice, which is when a court grants temporary admission to its bar to argue a particular case. LexisNexis indicates that motions for admission pro hac vice were far more common up through the 1970s—that means more everyday lawyers, few elite appellate litigators.

Compare this with the D.C. Circuit, whose bar professionalized much sooner. This was probably for many reasons: statutes gave the Circuit preference in agency litigation, and agencies wanted to argue on their “home turf” anyways; the capital drew ambitious political lawyers into its fold; and the list goes on. The Circuit did become notable for its expertise in administrative law, but just as importantly, so did the litigators practicing before it.

Therefore, the D.C. Circuit overtook the Supreme Court in developing administrative law. Despite SCOTUS nominally having the final say on all questions of agency power, it could only take so many cases onto its docket a year. It required a gatekeeper, and that gatekeeper was the D.C. Circuit. While SCOTUS and the Circuit often butted heads—it was overruled on numerous landmark cases—yet it silently kept up its work of reforming crucial doctrines, with the understanding that its bar was more competent to speak on such issues.

This confluence of factors—some institutional, some based on elite competence—led the D.C. Circuit to its role as preeminent overseer of the administrative state. Even when it rebelled against SCOTUS, it could often have the last laugh. As Scalia wrote after the Circuit’s tacit disobedience of Vermont Yankee v. NRDC: “the D.C. Circuit is something of a resident manager, and the Supreme Court an absentee landlord.” Sure, SCOTUS might grant certiorari, but more likely, the Circuit would have the final say. Thus, within the broad strokes painted by SCOTUS, the D.C. Circuit silently filled the interstices for decades, often with great sympathy to bureaucracy.

Revolt on the Frontier: State Solicitors General

With the Beltway’s well-organized bar and home-turf advantage, the inertia in favor of the administrative state seemed insurmountable. Major conservative public-interest groups like the Institute for Justice and the Center for Individual Rights didn’t spring up until roughly the early 90’s, by which point the Circuit’s institutional prestige was firmly entrenched. How then did they manage to successfully whittle away at the Circuit’s doctrines, culminating in the Supreme Court recently overturning Chevron?

The conservative lawyers avoided the already-hijacked federal apparatus and turned instead to state governments. Cue the state solicitors general: lawyers in state departments of justice dedicated solely to representing state interests at federal appellate courts such as the Supreme Court—interests that were, crucially, often anti-administration.

As Dan Schweitzer noted, state SG offices exploded in popularity at the turn of the twenty-first century. In the 1980s, only eight states had SGs. Today, forty-four states have one. Attending the expansion in state SG offices is a greater involvement of states in constitutional litigation more generally. As Will Baude and Sam Bray write: “[i]n just the last decade and a half, states have come to dominate the public law scene,” adding that today’s preponderance of state plaintiffs was not typical of earlier generations.

The formation of these offices pooled resources and talent that defended (often conservative) states’ prerogatives. As a result of this, the quality of legal argument at SCOTUS has improved substantially. Compare the desultory historical analysis of legal opinions from the twentieth century—often a few stray quotes from The Federalist would suffice—to the sophisticated briefs filed today. And these sophisticated briefs often aimed to roll back the worst excesses in administrative law.

The fruits of this long structural overhaul have finally come in the form of landmark, state-led cases to overturn long-standing administrative law. The chinks in Chevron’s armor began as early as 2015 in Michigan v. EPA, where, at the state of Michigan’s urging, the court tightened up how “reasonable” an interpretation had to be to receive deference. The opinion was written by Justice Scalia, who was himself instrumental in expanding Chevron as a judge on the D.C. Circuit. Later in another state-led case, West Virginia v. EPA, the court narrowed Chevron so much as to be effectively a nullity. The decision was finally overturned in Loper Bright v. Raimondo, which featured an amicus brief by eighteen state solicitors general.

Other challenges are just on the horizon. The current case about nationwide injunctions, Trump v. CASA, came up to the court as a consolidation with two other state challenges, Trump v. New Jersey and Trump v. Washington. Here, Tennessee’s Solicitor General wrote a brief which argued that nationwide injunctions violate Article III. If the Supreme Court does away with universal relief, this would signal yet another win of conservative states over the D.C. Circuit, who pioneered universal relief all the way back in the 80s.

In short, the struggle over defining the proper limits of administration offers insight into what drives legal reform. The D.C. Circuit captured high-quality legal talent, and won its reputation and power thereby. But a disaffected counter-elite accumulated over time, organized its ranks within the state solicitors general, and slowly but surely chipped away at leviathan’s armor.

To the competent go the spoils.

I’ve worked on this piece for years.

The project kicked off when, quite some time ago, I discovered a remarkable fact: Congress had received better oversight information on government spending in the 1930s than it does today. It received literally hundreds of pages of reports itemizing spending in every possible way – much more elaborate information than Congress gets now. And Congress received these reports weekly. To top it off, this experience had apparently never been discussed in any secondary literature.

This discovery raised two main questions. How could this possibly have been true? And if Congress had received better information 90 years ago, what the heck changed?

I looked into it. American Affairs published what I learned. Here’s some of it:

The introduction

The debates over the Department of Government Efficiency have revealed, if nothing else, that the federal budget is obscure even to the political combatants ostensibly responsible for developing and overseeing it. In the executive branch, Elon Musk highlights that billions of dollars of payments are processed by the Treasury without even a memo line. Meanwhile, in Congress, Republican politicians highlight the incompleteness of the bureaucracy’s spending records, while Democrats bemoan the Trump administration’s dissimulation in ceasing to share budgetary guidance documents. The camp followers of these obscure programs are thousands of federal contractors, pursuing vague goals with indefinite timelines. As soon as the ink on a bill is dry, it seems, Congress loses sight of its initiatives until their eventual success or their all-too-frequent failure.

Contrast this with the 1930s, when the Roosevelt administration provided Congress with hundreds of pages of spending reports every ten days, outlining how tax dollars were being put to use in minute detail. The speed and thoroughness with which these reports were produced is hard to fathom, and yet the administration was actually holding its best information back. FDR’s Treasury had itemized information on hundreds of thousands of projects, down to the individual checks that were written. Incredibly, politicians had better dashboards in the era of punch cards than we have in the era of AI. The decline in government competence runs deeper than our inability to match the speed and economy of New Deal construction: even their accounting was better. What happened? […]

The 1930s: Accounting Rebuilds America

In the 1930s, the United States faced a once-in-a-lifetime cataclysm. After the stock market collapse in 1929, unemployment rose to 25 percent. Cities and states were functionally bankrupt. FDR’s landslide victory against Herbert Hoover gave his administration a mandate for aggressive reform and experimentation. The administration proposed unprecedented spending for economic recovery, and from 1935 to 1937 was granted the then stunning amount of $8.5 billion for infrastructure investment alone—roughly $200 billion in today’s dollars.

This New Deal spending built roads and bridges, while the need to account for the spending spree built state capacity. To keep track of these staggering sums, the Treasury Department created an accounting system that provided nearly real-time information on more than two hundred thousand individual projects, down to recording every last check within ten days of it being written. This information was shared with both the White House and Congress, who invested in the bureaucracy due to the trust this information generated. Achieving this unprecedentedly good oversight required teamwork in the form of an alliance between Treasury Department bureaucrats and New Dealer academics. The Treasury bureaucrats were interested in better accounting for its own sake—it was their job, and they wanted to do it well. For FDR’s academic advisers, however, these reforms were a means to an end: increasing executive power. […]

The 1950s: Revenge of the Gray Flannel Suits

To 1950s businessmen, the dream of rolling back the size of government seemed like a lost cause. The federal government undertook a range of responsibilities that would have been unthinkable before the Great Depression and which had become unchallengeable after war’s end. Democrats had won nearly uninterrupted control of Congress since 1932, and the Republicans only won the Presidency by nominating the war hero Dwight D. Eisenhower, who professed his acceptance of the New Deal.

Conservatives had lost the hearts of the voters but hoped they might yet win the hearts of government accountants. Their suggested reform was program budgeting, which aimed to measure the total cost of government programs more comprehensively than before. This ap­proach won the endorsement of the landmark Hoover Commissions and was ultimately enacted in law. These reforms, however, inadvertently decreased state capacity. To measure costs in a broader way, agencies had to expand the role for budget offices at the expense of technical experts. This cost data was not useful or even desirable to the politicians in Congress who oversaw agencies. It did appeal to its backers, however: an alliance of accountants and businessmen. […]

The 1960s: The Slide Rule Reformers

Liberals of the 1960s were ready to break with the stodgy conservatism of the Eisenhower era. From closing the missile gap, to improving national education, to dealing with southern racism, the U.S. political system they took control of appeared unable to confront issues of dire importance. In one significant area, however, liberals doubled down on conservative reforms: the accounting and budgeting procedures with which they intended to fund a more vigorous government.

The government was confronting a daunting set of crises, and in their view, it was time to get down to brass tacks. The bureaucracy had to escape the squabbling of politics and figure out what to tackle head-on. It had to set its long-run goals and ruthlessly prioritize its programs to achieve them, an approach they called Planning, Programming, Budgeting (PPB). Robert McNamara, the secretary of defense in the Kennedy and Johnson administrations, pioneered this approach in the Department of Defense, eventually persuading Johnson to mandate it governmentwide. Unfortunately, far from achieving reformers’ ambi­tious goals, the approach dealt a lasting blow to government competence: PPB effectively centralized power with the planners and froze out technical experts, eventually leading the government to farm out policy to contractors and NGOs. The ultimate results were the last thing in the world its supporters—businessmen and the military’s social scientists—would have wanted. […]

What else, you ask?

Beyond these excerpts, the piece also includes such key details as: an overlooked public interest film, questionably qualified urban reformers, a statute with a suspicious job description, the annoyingly complicated way that items from the National Archives are cited, cattle research, one of GSA’s many scandals, military contractors moonlighting as education consultants, the vagaries of the Coast Guard’s budget, and why DOGE ought to learn three things and also have a movie night.

Read the rest of Accounting for State Capacity at American Affairs.

Previously I’ve written about the government’s most striking successes. It’s now time for a series on government failures and what they can teach us. And not one-off failures, either. This post introduces a new series on efficient executive branch reorganization – approaches for reorganizing the government to address the most stubborn problems in public policy.

So what are the worst of the worst in government policy, these evergreen problems? Herbert Emmerich has an answer.

Herbert Emmerich was a longtime civil servant and academic. He ran several agencies during the 1930s, eventually becoming executive secretary of the War Production Board during World War II. He was chairman of the Public Administration Service (an association of public administrators), and consulted for international organizations such as the United Nations. In short, he saw it all from every angle. In his last year of life, he wrote the book Federal Organization and Administrative Management, which summed up his lifelong thoughts on public administration.

His chapter on the organization of the executive branch discussed what he considered to be the thorniest issues. In particular, he highlighted five agency structures and types of programs as being uniquely and consistently problematic:1

1. Independent regulatory agencies, such as the Securities and Exchange Commission (SEC) or the Federal Communication Commission (FCC).

2. Government-owned corporations, such as Fannie Mae or the Tennessee Valley Authority (TVA).

3. Intergovernmental programs, including aid to states and programs that are jointly administered by federal and state government (such as disability claims or Medicaid).

4. Programs administered by contractors. These include everything from Head Start, whose programs are run by nonprofits, to military research, which is often conduction by nonprofits such as RAND.

5. Programs requiring special external coordination. For instance, national security is coordinated in the White House by the National Security Council, which ensures that diplomats, spies, and military officers are all marching to the beat of the same drum.

Old hands in Washington would agree that these programs cause no end of headaches. But why is this the case?

Schoolhouse Rock and its discontents

To understand what makes these areas problematic, consider what makes ordinary government programs unproblematic.

High school civics class teaches how a typical and unproblematic government program works. In this “Schoolhouse Rock” theory of public administration, a government program begins with Congress identifying an issue that is a federal responsibility. It then passes a law that sets a specific goal and empowers an agency to accomplish it. The President then picks the people to run the program, who implement the goal through directing the federal bureaucracy.

Meanwhile, the other two branches of government provide oversight. Congress holds hearings and wields its greatest influence through its “power of the purse”, the ability to cut off funds to misbehaving agencies. The judiciary stands above the fray and polices government actions, thereby ensuring that the constitutional and other legal rights of private citizens are protected.

Experienced observers could find at least a dozen problems with this oversimplified story. Nonetheless, for many government programs it is more or less an accurate description.

This story is more than a theory of civics – it is a theory of management. It starts with some specific task being identified from a well-defined set of federal issues, which is then made concrete by a carefully drafted act of Congress. Once the goal is identified, the program is carried out via a chain of command running from the President to a cabinet secretary to whoever is managing the specific program. External oversight is provided by Congress, which reviews the overall adequacy of implementation (somewhat akin to a corporation’s board of directors), while the judiciary reviews the legality of policy (somewhat akin to a corporation’s external auditor).

But by contrast, the five areas above are uniquely problematic because they are the least similar to the Schoolhouse Rock theory of government. They deviate the most from this theory of management and yet they must be managed, leading to intractable arguments over the form management should take instead. Consider each of the five areas in turn and how they deviate from a naive theory of government.

The five problem children

Regulatory agencies – Regulatory agencies set the rules of the game for particular industries. For example, the Federal Communications Commission (FCC) regulates cell phones, radios, and other uses of the radio spectrum. Regulatory agencies typically have a range of responsibilities in order to carry out this mission.

Primarily, regulatory agencies typically write general rules for their industry – for instance, FCC sets rules about which parts of the radio spectrum are reserved for radios vs for cell towers. Generally, regulatory agencies review and approve permits – for instance, FCC reviews applications from people who want an amateur radio operator’s license. Regulatory agencies adjudicate complaints against the companies they regulate – for instance, FCC takes action against phone companies that are responsible for spam phone calls. Additionally, regulatory agencies sometimes administer various grant programs or have other assorted tasks.

These regulatory agencies tend to be structured as independent commissions who vote on the rules and the adjudications. That is, a bipartisan commission of about five people run the agency, which is not part of a larger cabinet department. The commissioners are appointed for an extended period of time. This structure raises several issues.

• The agency’s independence weakens presidential oversight. The president picks a bipartisan board to run the agency, rather than choosing a trusted lieutenant to have full authority to run it. Not only does the President have less influence, but there is less clear of a chain of command overall in this multi-headed agency.

• The agency’s regulatory actions are less subject to judicial oversight than other government actions. When agencies adjudicate complaints, they often have procedures that are similar to courts, yet lack independent judges. For instance, if FCC wants to revoke the license of a radio station, it has a trial-like procedure where the radio station’s lawyer can present evidence. The civil servant running this procedure is not a judge, but an employee of FCC (the agency attempting to take away the license). It is an disputed question if these procedures are an acceptable substitute for real trials.

• Finally, regulatory agencies often (but not always) fund themselves with fees, which weakens Congressional control. Agencies charge the public fees to, e.g., file applications, which makes them less dependent on Congressional appropriations. This, in turn, weakens Congress’ power of the purse.

Government corporations – Government corporations are agencies whose work is similar to the work of private corporations. For instance, they might originate loans, offer insurance, or provide services to consumers. Federal examples might include: the Tennessee Valley Authority (TVA), which sells electricity in the Southeast much as any electricity generator would; Fannie Mae and Freddie Mac, which buy mortgages from banks and package them into securities; and the Federal Deposit Insurance Corporation (FDIC), which charges banks fees to offer them deposit insurance, thereby insuring that an average person does not lose money if a bank goes bankrupt. States and localities often have government owned water or electric facilities.

Government corporations often have substantial autonomy and are generally not part of cabinet departments. They are generally run by a board, which sometimes picks the head of the agency (although the President sometimes picks the head). Often, nearly their entire funding comes from their fees and other revenue. Once again, issues arise from the organization of these agencies.

• Autonomous government corporations often lack a clear chain of command, as they are rarely part of larger cabinet agencies and (sometimes) the president does not pick their director.

• Government corporations receive revenue from their operations, which greatly weakens Congressional oversight. For instance, TVA gets its funding through selling electricity, which means that Congress has relatively little influence over it (compared to other agencies).

Intergovernmental programs – Intergovernmental programs are those that are jointly administered between the federal government and state/local government. Often these are aid programs where, for example, the federal government might aid states that are already building roads by offering to cover part of the cost. In other cases, federal programs are formally administered by the states. For instance, Medicaid is a federal program that provides health insurance to the poor. The federal government sets Medicaid eligibility and coverage standards and provides funding, but the states design and implement the programs.

Here, the common problems arise from the nature of these programs rather than the structure of agencies.

• Intergovernmental programs have a weak chain of command and weak oversight across the board. When the states and the federal government are both in charge of an issue, then nobody is truly in charge of it. For instance, when the federal and state governments are jointly funding highways or other transit programs, both parties are more willing to be wasteful because they are wasting other people’s money. When intergovernmental programs fail, Congress and the president lack a unique person or agency to hold accountable.

• These divided responsibilities not only create problems of oversight but also weaken the system of federalism. The more cooperation there is between the federal and local governments, the muddier is the concept of unique, clear roles for each layer of government.

Programs administered by contractors – Contractor administered programs are programs that the federal government farms out to a third party rather than performing itself directly. Sometimes this entails outsourcing simple office functions, such as when agencies pay contractors to update IT systems rather than using their own IT team. Sometimes the purposes are broader as, for example, how the military contracts with RAND to serve as an official think-tank. In some cases, the role of contractors is extraordinarily expansive and they administer nearly the whole of a government program. For example, the FCC administers a program called the Universal Service Fund. Rather than administer this program itself, FCC pays a nonprofit – namely the Universal Services Administrative Corporation (USAC) – to carry out the work.

As above, problems arise from the nature of the program.

• Contractual programs can lack any sort of chain of command and almost entirely escape oversight. Contractors are not part of the bureaucracy at all and are therefore minimally responsible to the president, Congress, or the judiciary.

Programs requiring special external coordination – In some cases, many agencies might be involved in the same policy area, thereby requiring external coordination. For example, the National Security Council (NSC) coordinates the nation’s spies, diplomats, and military officers to ensure they have a consistent vision of foreign policy (namely, the President’s). Similarly, the legal policies of government agencies are coordinated by the Department of Justice so that the government can speak with one voice – it would be somewhat absurd if two agencies were arguing against each other in court.

• Here, the problems arise because interagency cooperation is difficult. Agencies would always rather have the freedom to push their own viewpoint and dislike being coordinated by an external entity.

Efficient reorganization: a series

This post introduces a new series on efficient executive branch reorganization. These five areas are the most intractable problems, so efficient ways of reworking them offers the greatest value. How can these programs and agencies be restructured or reorganized to minimize the problems mentioned above? How can these corner cases be made to conform as closely as possible to the Schoolhouse Rock theory of government?

Although executive branch reorganization is in the news (due to DOGE), most of the current proposals – cutting headcount and eliminating minor agencies – would make government organization only marginally more efficient, at best. This series aims, by contrast, to get at the heart of the matter.

Each type of program or agency will receive one or more posts. I’ll lay out the history of the problem, the history of proposals for reform, and the (not too extensive) theory of public administration needed to make sense of it all. I’ll sum up by indicating what reorganization would look like today if the goal is to bring these programs/agencies closer to the Schoolhouse Rock view of government.

Even if these problems can never be solved, the lessons of history and administrative theory show us how they might be improved.

If you asked DC insiders to define red tape, they would point you to the Paperwork Reduction Act (PRA). This widely-loathed law regulates the process by which the government may request information from the public.

The PRA attempts to prevent government agencies from asking too many duplicative questions to the public. The Act applies if an agency wants to ask a question to more than ten people. If so, the agency must go through a lengthy approval process (normally taking many months), which involves 1) review by the staff at the President’s Office of Management and Budget, and 2) opportunity for public comment.

This requirement makes it practically impossible for agencies to conduct user research or ask the public if government programs are working. Absurdly enough, the public-comment requirement makes it mandatory to ask if the government ought to ask a question, while also making it difficult to actually go ask the question! The PRA is intended to prevent the government from harassing citizens, but in effect prevents government from improving its processes.

Others have ably made the case against the (perhaps comically-misnamed) PRA, resulting in some improvements to how OMB approves user research. And reformers are continuing to press the case – I’ve even heard rumors that there will be several other must-reads on PRA reform published today, coordinated by the Niskanen Center. Keep your eyes peeled for pieces by Marina Nitze, Jennifer Pahlka, and Alex Mechanick!

But even with all this work, another question remains to be answered. Excessive paperwork is a perennial issue, so how did the government approach this problem in the past?

In the mid-1940s, paperwork was a dire problem. Since the New Deal, government agencies had been created at a furious pace, with each new agency inventing its own approach to paperwork. As the war ended, the public was sick of the confusing and user-hostile paperwork that was imposed on it.

The federal government was determined to get this under control. As part of broader management reforms, the Bureau of the Budget invented an approach to reducing the burden of paperwork called forms control. They grappled with the same problem that we face today – was their approach any more popular than the PRA?

In fact, forms control was so good that even private businesses begged the federal government to teach them its approach.

Forms control

Forms control was an approach to managing office paperwork, including both forms that the public filled out and (unlike today) purely internal office paperwork. That is, it was a process for an agency to keep track of both the number and substance of every form whatsoever that it used.

The Bureau of the Budget had begun researching ways to manage paperwork during World War II, inquiring about the best practices of both businesses and government agencies. With this research, it produced a manual, Simplifying Procedures through Forms Control, which taught a new systematic method for applying these best practices. (All quotes and graphics in later sections are taken from this manual.)

The manual stated that forms control had four goals, which (paraphrased) were:1

1. Eliminating unnecessary forms.

2. Establishing a standardized design language for forms, thereby ensuring uniformity and simplicity.

3. Producing and distributing forms as economically as possible.

4. Using the process for creating and revising forms to address underlying problems of office organization and procedure.

To accomplish this, agencies designated a small centralized team to manage the entire agency’s paperwork. Other teams who wished to create a new form or alter an existing one would file a request with this team. From here, the (paraphrased) steps to apply forms control were:2

1. Recording the request to create or revise the form.

2. Analyzing the form’s purpose and its relation to internal office procedures.

3. Applying uniform design standards to the form.

4. Assigning an identification number to the form.

5. Approving a final specification for the form.

6. Planning how to store and distribute the resulting form.

7. Evaluating the effectiveness of the forms control program itself.

The manual went on to explain each step. The manual received an unusual honor for government office manuals: it was reviewed by an academic journal, which stated:

The bulletin Simplifying Procedures through Forms Control is very nearly a definitive work on the subject. Although it was originally prepared solely for the use of federal agencies, so many requests for it were received from business firms, state and municipal governments, and universities, that it has now been reprinted for wider distribution.3

Of course, being a bestseller among businesses was even more unusual of an honor. Why was forms control an inspiration to the private sector, when the PRA today is among the most dysfunctional government procedures? For a start, forms control was done at each individual agency, with the Bureau of the Budget mainly offering advice rather than mandates.

But most importantly, forms control was tied to both user experience and procedural improvement.

Better user experience

The Bureau focused on user experience through creating a standard design language for public forms. It included everything from approved fonts, to standardized sizes of paper, to a style manual for words and phrasing. The excerpts below are taken from the Bureau of the Budget manual:4

Fonts:

Style manual:

The Bureau then pushed government agencies to revise their forms to conform with this design language.

If you’ve ever noticed that federal government forms all look the same, it’s because the Bureau of the Budget pushed hard to standardize them in the 1940s. If you’ve never noticed, it’s because the Bureau of the Budget succeeded!

The guide shows a “before” and “after” of an actual government form to illustrate the design language being applied. The “before” form looks awful – e.g., it is difficult to see how many fields need to be filled out. The “after” looks like modern government paperwork.

Before:

And after:

How did the Bureau create this design language? While user research was primitive in the 1940s, the Bureau noted that their design language was the product of collaboration with both other government agencies and relevant experts from civil society (such as librarians and publishers).

Today, the PRA is administered by bureaucrats with little knowledge of design or user research – but in the 1940s, the government’s practices were cutting-edge.

Better procedure

Forms control was also tied to procedural improvements. Agencies went through an approval process to create new forms, but also (unlike today) periodically reviewed existing forms in order to streamline and improve the way that an agency did business. To accomplish this, the agency would take a copy of every form and file it in a functional file.

The functional file (above) contained copies of forms grouped by major purposes. It might, for instance, have a group called farm aid: this might includes a form that the public filled out called application for aid and also include a form that only the agency filled out called review of aid application.

This file would be used to drive operational improvements. Managers were required to periodically look through related forms to see if duplicative forms might be combined. But more than this, the functional file showed public-facing paperwork side-by-side with purely internal paperwork. Managers were required to analyze ways to improve their office procedure: they might improve their use of the information that the public submitted, and conversely, they might stop asking for information that the agency did not actually intend to use.

Forms control was intended to serve the agency’s goals of streamlining its internal procedure in order to deliver high quality service, rather than be an end in itself. It was even supposed to speed up government rather than be red tape; indeed, the real purpose was to draw attention to the agency’s underlying procedures to improve faulty processes. The concluding section of the guide clearly explains this point:

In general, however, forms are a reflection of the work methods, operating procedures, and management know-how which give rise to their use. If an agency’s forms constitute a simple, orderly plan showing clear and related purposes, there is reason to believe that its personnel–knowing what they are doing and why–may be giving fairly efficient service. If, on the other hand, its forms constitute an unintelligible tangle of red tape, it is pretty safe to assume that its methods and procedures–its service to the public–are in much the same shape.

Therefore, great as may be the savings from eliminating unnecessary forms and getting all necessary ones properly related to existing procedures, forms control offers a still greater value as a means of detecting the need to treat the procedures themselves.

Paperwork: then and today

Forms control was part of the Bureau of the Budget’s broader set of management improvement initiatives, which went hand-in-hand to build up government competence. I have described two others: process charting and plain language. But what can we learn from forms control, specifically? Two things in particular: a better vision for government procedure, and a sense of how high reformers ought to aim.

First, it offers a better alternative to the Paperwork Reduction Act. Today, the PRA is administered by the Office of Management and Budget, or more specifically at its Office of Information and Regulatory Analysis. The small and inexpert team is a significant bottleneck on paperwork approvals that fails to add much value. More than that, the approach is purely a mandate. The PRA increases the burden on the government in hopes of indirectly benefiting the public.

Contrast this with the 1940s, where the forms control initiative invested in agencies, training them to minimize the burden on the public. This (mainly voluntary) guidance taught agencies to focus on user experience and operational improvement in order to better serve the public.

We could imagine a similar approach today. As a rough sketch, requests for information might be a chance to force collaboration between IT teams, data teams, and user experience experts. When an agency requests information from the public, it might review 1) the user experience the public faces in supplying the information, 2) the IT tools the agency uses to handle this information, and 3) how the requested information relates to other data the agency is already collecting. The US could, once again, view requests for information as a chance to improve the way that agencies do business. More broadly, we could abandon punitive approaches and figure out how to invest in agencies today.

Second, forms control shows us how much government can achieve when it buckles down. The approach of forms control was so successful that businesses and universities demanded that the government reprint a dry office manual. The Bureau of the Budget definitively answered the problems of the paper-based era and anticipated many issues of the digital era.

That is how high reformers ought to aim.

Appendix: Sources, and a question for readers

You can read the forms control manual on Google Books. It is excellent – note that I describe only two aspects of a broader program. My discussion leaves out many steps that are not directly relevant to UX or operations.

Sources:

• Reilley, Ewing W., and Richard F. Neuschel. “Administrative Management Know-How.” Public Administration Review 10, no. 4 (1950): 291. https://doi.org/10.2307/972771.

• US Bureau of the Budget. “Simplifying Procedures through Forms Control.” US Government Printing Office, June 1948.

Question for readers: When did your government improve its forms? This question applies both to foreign governments and to state and local government. My friend works at a state government and said they still used old-fashioned forms until the 1990s. (He claims that IRS-style fill-in-the-box forms were “an invention on par with the abacus, if not slightly more important.”)

I would love to know the foreign history and how the US compares.

Subscribe now

After years of government failures, an angry Congress confronted the federal bureaucracy. The much-criticized bureaucracy used archaic technology and was badly behind on its business. Many employees didn’t even bother to show up to work in person. Congress and the nation were ready for a thorough overhaul of the executive branch. The year was 1887.

Just as the late Gilded Age bureaucracy was failing to adapt to the industrial age, today’s bureaucracy is proving itself inadequate for the digital age. There is once again a push for fundamental reform, embodied today in Elon Musk’s Department of Government Efficiency (DOGE). Lawmakers who support DOGE should draw lessons from the 1880s, when congressional investigations helped drive reform by uncovering bureaucratic failure.In the 1800s, accusations of government inefficiency and corruption were as common as today. So were congressional investigations of these allegations. Congress, in addition to routine oversight, had conducted at least 63 one-off investigations of particular issues. But all of these investigations relied upon information that bureaucrats voluntarily submitted, so nobody knew what the bureaucracy was hiding.

Although complaints from constituents indicated that delays and mistakes were more serious than agencies admitted, nobody knew the depth or causes of the bureaucracy’s problems. Businessmen felt that agencies were hidebound and that civil servants were lazy, whereas officials felt that the government was inadequately staffed. In 1887, Senator Francis Cockrell hit upon a novel solution: He formed what came to be called the “Cockrell Committee,” whose members set out to see the bureaucracy’s problems with their own eyes.

Read more of my piece The DOGE of the 1880s over at The American Conservative. Over the next few weeks, I’ll go into greater detail about this significant and largely forgotten reform initiative. Until then, enjoy the article!

The second World War called for national sacrifice: hundreds of thousands of soldiers gave their lives, while millions of civilians ate rationed food and skimped on goods needed for the war (such as rubber and paper).

Total war also involved less traditional sorts of sacrifice, with citizens donating their goods directly to the military – sometimes even donating their pets.

Most of these donations were requested by the federal government. However, many patriotic Americans sprang into action and donated anything they felt the military ought to have. They didn’t wait on mere formalities, such as finding out if the government actually wanted it.

Donations during the war

There were a wide variety of donation drives throughout WWII. These included campaigns for rubber, paper, kitchen fats, women’s stockings, and numerous other goods. The military asked for specific items: for instance, the Navy asked for high-quality binoculars.

Even pets were enlisted into the war effort. In the Dogs for Defense program, the military asked families to send their dogs to serve as sentries (and similar roles).

Besides all of these formal drives, Americans felt free to donate other goods, whether the government had asked for them or not.

And we have a very entertaining description of what they donated. The US Department of the Treasury kept records on all donations, both solicited and unsolicited. In 1946, a Treasury official named E.F. Bartelt reviewed these records and said (of the donations):

They included motor boats and yachts; motion picture films; a train of eight coaches; decontamination units; around two million copies of various books, booklets, and magazines; ten million yeast tablets; thirty thousand deep-sea fishing kits valued at $65,000; twenty-five thousand game kits valued at $82,000; baseball kits valued at $22,000; medical supplies, including iron lungs; public address systems for hospitals; radio equipment; a completely furnished day-nursery building; a greenhouse; eighty thousand phonograph records; more than fifteen thousand carrier pigeons for the Signal Corps; several hundred dogs for war service; cigarettes valued at $153,000; a riding horse; also many bed furnishings and furniture; handicraft equipment; nursery supplies; thousands of decks of playing cards; flags; meal tickets, and many other items.1

But this was no one-sided bargain. The Treasury department sent a certificate of appreciation to everyone who donated their worldly possessions, whether the government had wanted them or not:2

It was unquestionably a different era.

Appendix: The author wants YOU

I would love to know more about the train, the completely furnished daycare, the riding horse, and the greenhouse.

Do you know anything about these intriguing donations? Or know somebody who would know?

If so, I’d love to hear more. Shoot me an email at khawickhorst@gmail.com, or message me here on Substack or on LinkedIn.

Some few government agencies have a bias towards doing things – what Caleb Watney calls the operational mindset. The Census Bureau, for instance, exists to take a census. Most agencies do not, which is often a consequence of the way they are organized.

In Agency Missions, I discuss two ways to organize agencies and the typical mindset that results from each. One way to organize agencies is by subject matter – for example, the Forest Service does everything related to the subject of forests (including functions such as research and administration). Another is by function – for example, the National Science Foundation undertakes the abstract function of research (and researches many subjects). A subject matter agency has a concrete mission, and therefore tends to have an operational mindset. A functional agency tends to find this mindset unnatural.

In The High Tide of Reform I use the US Department of Agriculture (USDA) as a case study of these two types of agencies. Its 1942 reorganization turned its subject-matter agencies into functional agencies, which was almost immediately followed by the department losing its political entrepreneurialism and being captured by the agricultural interest groups. The reorganization demonstrates the harm of removing an operational mission.

But the history supports an even stronger claim: the decline of USDA’s operational mindset was not only a consequence of functional reorganization – it was an intended consequence. The farm lobby had strongly endorsed functional reorganization in order to make the USDA bureaus more blindly responsive to the farmers lobby, and therefore less mission-oriented.

So does an operational mindset lead to greater state capacity? Almost certainly. Even the opponents of state capacity agreed – namely the farm lobby. The farm lobby debated USDA organization using these concepts, and eventually won control over the previously-independent bureaucracy.

The lobby’s ultimate success in capturing the bureaucracy didn’t happen overnight. Their victory in the 1940s was laid much earlier, through their planning in the 1920s.

Live at farm aid, 1925

For American farmers, the 1920’s were a troubled decade, marked by turmoil in a time of prosperity. In 1925, President Coolidge responded by convening a farm conference, which studied the problems facing American agriculture and suggested ways to aid farmers. Among much else, they considered the question of departmental organization at USDA.

Their discussion is so instructive that I quote it at length:

The activities of many different departments and agencies of the Federal Government have a direct bearing upon agricultural welfare. In general, these activities may be divided into two major types: namely, service functions and regulatory or law-enforcement functions. Service activities consist essentially in the accumulation and dissemination of information concerning all factors which enter into the production, distributing, and consumption of agricultural products, and advice and assistance in putting this information into practice. Regulatory functions consist essentially in the interpretation and enforcement of laws and regulations designed to protect the interests of both the producers and the consumers of agricultural products.

In many of the Federal departments, both the service and regulatory functions dealing with the same commodity or industry are lodged in the same bureau, offices or personnel. This has many disastrous effects. In the discharge of the regulatory or police functions, officials of the departments are sometimes required to adopt the judicial attitude, sometimes a combined judicial and prosecutorial attitude, but more often an exclusively prosecutory. This attitude inevitably leads to a feeling of antagonism of interest between the department officials and the individual citizens or organizations which come into contact with the Federal agency. Such a feeling is the exact opposite of that which must maintain if the service functions of the agency, which depend upon a community of interest in advancing the welfare of the industry, are to be effectively discharged. Many of the instances of unsatisfactory administration of government activities touching agricultural welfare, which have been brought to the attention of the conference, have been clearly and directly traceable to the feeling of antagonism, instead of community of interest, between the government officials and the individual or group which was seeking government assistance.

The Conference, therefore, recommends that in all branches of the Government, the service functions and the regulatory functions be separated as completely as possible in organization, personnel, and action.1

In some respects, this report was nothing new: farmers had (naturally enough) advocated for a more pro-farmer USDA since its creation. But this presidential conference was, so far as I know, the first chance for the farm lobby to think through what a favorable reorganization would look like. I would tentatively say that it was even the first major discussion of how departments should, in principle, be reorganized.2

Regardless of if the report directly influenced the 1942 reorganization, it was an early statement that the farmer’s lobby would support functional reorganization, and was unique in explaining their reasoning at length.

Reorganization and its consequences

The passage shows that the idea of “functional organization” or “operational mindsets” aren’t modern ideas grafted anachronistically onto the past: they were precisely the concepts that interest groups used to discuss bureaucracy, even in the 1920s.

The farm conference rejected subject matter agencies that combined service (i.e., grants and education) with regulation and research. For instance, the Bureau of Entomology researched insects, condemned diseased crops, and educated farmers on preventing infestations. The farm conference felt that these were incompatible goals: the farmers would be far more positive about the agency if it only gave them grants and education, without also regulating them.

Accordingly, they advocated that regulatory and service functions be divorced. The only logical way to reorganize the department (under their scheme) would be to group regulatory agencies together and service agencies together; that is, to have a some number of agencies devoting to regulating farmers, and some number devoted to distributing aid. Through this functional reorganization, the agencies that distributed aid would form a “community of interest” with the farmers.

Now, if the goal is a “community of interest” with the farmers, it follows that the goal is not some operational mission. The opponents of functional reorganization made this even more explicit, as in Gifford Pinchot’s argument against similar (proposed) reorganizations:

If anything is proved in government work, it is that to separate administration and research means bad administration every time. As good an illustration as I know is the General Land Office in the Department of the Interior. Research it had none, and its mishandling of the public lands became a scandal, the stench of which is with us yet.3

Swap out “regulation” for “research”, and his argument holds exactly. Previously, a USDA agency had a mission, perhaps entomology. Its mix of regulation and service gave its regulators greater prestige, while also keeping the interest groups at a certain distance from the grant writers. But once USDA bureau lost their operational missions, they could only passively aid their interest groups; accordingly, their independence atrophied after the reorganization. As Pinchot would say, it meant bad administration.

Leading up to the reorganization, there was no debate about the effects. All involved believed that functional reorganization would lead to greater interest group influence and less agency entrepreneurialism. They only disagreed about if this was good or bad.

Eventually, the reorganizers got their wish, and the farm lobby got its desired influence shortly thereafter. But today we might turn that on its head – subject matter reorganization could help the government do things.

In popular imagination, the government spends its money by employing bureaucrats to harass an already much-harassed citizenry. Whatever the case may be, it is certainly true that salaries are a large expense for the federal government. So how does Congress decide exactly how much money will be spent on government salaries?

Or to be more precise: an agency needs Congress’ approval to hire workers. How many different ways can it receive this permission from Congress?

There are four fundamental methods.1

• Lump sum – an agency is given a total amount of money for salaries, with no further restrictions.

• Statutory definition – Congress creates a position and sets its compensation directly, in statute.

• Pay-grade allotment – Congress prescribes a standardized pay scale, which is not (necessarily) intended to strictly correspond with job responsibilities. Congress gives an agency the right to hire a certain number of people at certain pay-grades, but the agency writes job descriptions without anyone else being allowed to object.

• Position classification – Congress sets a standardized pay scale that is intended to correspond to job duties: if two workers do the same sort of work, they ought to have the same salary, regardless of what agency they work for. Accordingly, a single central personnel agency is tasked with reviewing all government job descriptions and classifying these positions, thereby setting their salary.

Each approach has a different tradeoff with respect to 1) legislative control of spending, 2) executive flexibility in administration, and 3) fairness toward government employees.

All four approaches were used at some point in US history, but position classification has been by far the most important approach ever since the Classification Act of 1923. The four methods are discussed in turn below.

Lump sum

The simplest way of controlling salaries is for Congress to appropriate a lump-sum of money for that purpose. For example, Congress might say that an agency gets $50 million to spend on salaries.

From there, the agency has total flexibility with respect to salaries – the only restriction is that it cannot spend a penny more than $50 million. For example: The agency can write any job description it wants and offer any salary that it chooses. It could offer two applicants precisely the same job, and then pay them entirely different salaries. The agency might increase or decrease salaries for any reason at all, and at any time. It implies total executive discretion.

This might sound fantastical, but it was how the US actually appropriated money for salaries up through about the 1830s. And the approach didn’t die out until much later – some agencies received lump-sum salary appropriations up to 1923.

By the time of the Classification Act 1923 (see below), this lump-sum approach was considered outdated. However, it had a minor renaissance in the 1990s – reformers aligned with the New Public Managment thought that its flexibility might foster entrepreneurial government. A National Academy of Public Administration report in the 1990s recommended experimenting with lump-sum salary appropriations; so far as I know, this suggestion was never followed.

Tradeoffs: This approach gives Congress total control over salaries. If Congress says an agency gets $50 million for salaries, then $50 million is exactly what will be spent. In turn, this gives the executive total control in using these funds – and therefore, the maximal possibility for misuse. It offers employees no protection against bias, as the agency can pay two identical employees entirely different salaries. (Of course, this doesn’t mean the agency will, only that it might.)

Statutory definition

An equally simple way for Congress to control salaries is for it to designate salaries in statute. The most obvious example is the President: Congress sets his salary in 3 USC § 102 – it happens to be $400,000.

Statutory definition was used throughout the 1800s and early 1900s and was applied more or less at random. Some positions’ salaries were set in statute but most were not; there was no particular reason for positions to be in the one group or in the other.

The Classification Act of 1923 replaced this approach.

Tradeoffs: In principle, this approach also gives Congress total control over salaries. However, amending statutes to change salaries is very time-consuming and therefore unlikely to happen – Congressional control was nominal instead of real. This approach has no executive flexibility. This approach also offers no assurance of fairness to government employees.

This approach is bad at all three potential goals: it is almost never the right approach, except for niche use-cases (such as for setting the President’s salary).

Pay-grade allotment

Another possibility is for Congress to define a standardized pay scale and allot an agency a certain number of positions at each grade of the pay-scale. The agency is responsible for drawing up job descriptions and assigning duties itself.

This can be illustrated with the actual US history. In 1854, Congress created a four-tiered pay-scale that lasted until 1923. There were four classes of clerks, each with their own salary: a class one clerk earned $900 per year; a class two clerk earned $1,200 per year; a class three clerk earned $1,500 per year; and a class four clerk earned $1,800 per year.

An appropriation might say, “The agency XYZ is permitted ten class one clerks and two class three clerks.”" Therefore, the agency would spend exactly $12,000 dollars on salaries per year. However, the agency could assign any sort of duties to its clerks – it might give the third class clerks less work than the first class clerks.

But in practice, a given agency typically did a logical enough job allocating work to the different classes of clerks. The real inequality was between agencies: some agencies thought that accounting (for example) was a job for class two clerks, while other agencies felt it was a job for class three clerks.

Pay-grade allotment was the typical approach from 1854-1923, until the Classification Act of 1923 abolished it.

Tradeoffs: This approach gives Congress complete control over spending and offers the executive branch reasonable flexibility. It is likely to be reasonably fair to government employees in practice, although in principle it might not be.

Position classification

The final possibility is to have a central agency analyze each job description and then set the corresponding salary. This approach prizes equality above all else, and aims to treat similar workers equally across the entire government.

In every previous approach, workers might receive unequal wages for equal work, particularly across agencies. One agency might treat all HR workers quite equally, and pay them all $50,000; another might treat all HR workers quite equally, and pay them all $55,000. But at the level of the federal government as a whole, HR workers are not being paid equally.

To combat this, an agency might be designated as the personnel agency – in the US, it is the Office of Personnel Management (OPM). This agency might be responsible for reviewing all job descriptions and analyzing how much skill is needed for their performance. With this knowledge, all jobs requiring similar levels of skill will be paid equally, which thereby creates a rationalized pay-scale. In the US, this pay-scale is called the General Schedule (GS) – OPM has rather elaborate guidance for classifying positions in the GS.

Naturally, putting a single agency in charge of salary decisions means taking it away from the rest of the executive branch. But less obviously, it also means taking it away from the legislature. For example, suppose that I, Kevin Hawickhorst, am a government employee with a lowly job title – Substack blogger, say. I might try to convince the personnel agency that my work is actually more technical and specialized than they had first thought – I am actually a historian of bureaucracy and a government reformer. If the personnel agency accepts my argument, then I get favorably reclassified and my salary is increased. The result: although Congress has not taken any action at all, total spending on salaries has increased.

Upward pressure on position classification is very common in practice, so this lack of legislative control is not merely theoretical.

Position classification was imposed by the Classification Act of 1923. This approach is much younger than the other three: the first position classification scheme (at any layer of government) was in Chicago in 1908.2

Tradeoffs: This option gives Congress the least amount of control over spending. It has minimal administrative flexibility and precludes several generally-desirable personnel practices.3 The only consideration is equal treatment of government employees – but even government employees frequently dislike the classification decisions of personnel agencies.

Personnel: then and now

Position authority is a very narrow issue: how Congress grants agencies the right to hire workers. This is separate from the question of hiring procedures – how specific applicants are interviewed and selected, and upon what basis; and separate from the question of discipline – how employees are separated from government service, and upon what basis. This also does not cover promotions, and therefore sidesteps the issue of pay raises for current employees.

But even for such a narrow issue, the US has adopted four wildly different approaches across its history. Ultimately, position classification has reigned supreme ever since the Classification Act of 1923 imposed it as the governmentwide standard. From there, the Classification Act of 1949 drastically expanded the scope of position classification, and its dominance has gone unquestioned since.

And this lack of questioning is surprising. The classification acts are bizarrely under-discussed given their importance and their dysfunction. These acts were controversial at the time of their adoption, and were consistently unpopular among federal workers.

More than that, there are alternatives: it is not a law of nature that the federal government must use position classification. There are three other approaches – and two of them are potentially good!

Appendix: Note on sources

My main source is Ismar Baruch’s book History of Position-classification and Salary Standardization in the Federal Service, 1789-1941. It can be described using several adjectives, one of which is “comprehensive”. On the off chance that the book’s title grabbed you, it is available on Google Books.

I have previously written about how the federal government once trained managers to focus on implementation.

• In my post Eisenhower’s Bureaucrats, I explained the approach to managerial training that the federal government developed, which they called work simplification.

• In my post The Snyder Cuts, I explained how Treasury Secretary Snyder put work simplification to practice at IRS: he modernized IRS’s IT while cracking down on corruption.

This raises a rather obvious question: if this approach was so good, why did the US abandon it?

As with all social trends, there were many reasons. So here I will note only one: competing theories about the purpose of government management.

Two theories of management

A surprising amount of government reform is the adoption of management fads. In this respect, the past was no different than the present.

During the Great Depression and second world war, the federal government’s administrators cast about for theories of management. Two ideas proved influential, and both were borrowed from private industry.

One was Luther Gulick’s POSDCORB idea, which was borrowed from French industrial management. The other was the Work Simplification method, which was borrowed from Training Within Industry (TWI), an industrial training program from WWII. The former focused on abstract policy from above, whereas the latter focused on bottom-up and iterative improvement.

The federal government eventually rejected TWI-based approaches in favor of policy planning from on high. Ironically, TWI ultimately developed into lean manufacturing in Japan; to contemporary eyes, the approach from the 1940s was more modern than what replaced it.

Gulick and POSDCORB

The ultimate victor in this struggle for federal management was Luther Gulick: an administrator, academic, and reformer. In his long 101 years of life, he achieved an enviable record – not least, he became a major figure in public administration despite having only written two or three article that anyone has read.

Among those articles was the immensely influential Notes on the Theory of Organization. This article poses the question, “What is the work of the chief executive? What does he do?” And according to Gulick, “The answer is POSDCORB”.1 This made-up acronym illustrates the major functions of the executive:

POSDCORB

• Planning

• Organizing

• Staffing

• Directing

• Co-ordinating

• Reporting

• Budgeting

Gulick adopts this idea from the managerial theories of Henri Fayol. Although Gulick is discussing the President, this is a general theory of management that could be applied at any level of any organization.

It is unnecessary to discuss the elements of POSDCORB, except that this managerial style is strikingly top-down and totally non-operational. There is no suggestion that the executive should understand anything at all about his organization’s work, much less that he should empower low-level workers.

These were intended as nearly the only issues for an executive to consider. Moreover, Gulick held that each function should be institutionalized: the executive should have a budgeting agency, a planning agency, a personnel agency, and so forth. These support agencies would help the executive make strategic decisions without the distraction of operational issues.

Gulick’s theory was immensely influential: he wrote the article as a background memo for FDR’s Brownlow Committee, which studied the executive branch and proposed reforms to increase its efficiency. Most of their proposals were rejected, but the proposals on executive management were largely adopted.2

The proposals were even more influential over the longer run. The reports were careful analyses, skillfully written by distinguished academics. The prestige of these reports (and of the presidency) ultimately affected management across the entire executive branch – cabinet secretaries and agency heads increasingly focused on these abstract goals. Departments created policy and planning units, and expanded their budgeting and personnel departments. The government became increasingly top-heavy and less focused on implementation.

But POSDCORB’s victory was not preordained. It won out against a competing theory of management, which was invented in World War II in response to the exigencies of total war.

Work Simplification and TWI

During WWII, the new war production plants needed to rapidly train line workers and managers. To assist with this problem, the federal government’s War Manpower Commission helped develop a standardized technique for industrial training, which they called Training Within Industry (TWI).

TWI covered many types of training, including managerial training. The TWI manuals focused on the lowest level managers – the supervisors – saying, “The supervisors are the key persons in the war production, home-front job.”3 The training then states that there are five main requirements for managers:

1. Knowledge of His Work.

2. Knowledge of Responsibilities.

3. Skill in Instructing.

4. Skill in Improving Methods.

5. Skill in Working with People.

This, too, could be used as a general theory of management. In fact, the federal government did use it as a general theory of management: Work Simplification was Training Within Industry as adapted for government managers.

There is once again no need to discuss the five requirements in detail, except that they focused on operations and implementation. Once again, this was a general theory of management – although the guides discussed the lowest level supervisors, it was a framework for managers at any level.

Of course, the top managers had to focus on other issues, too: executives undoubtably did budget, and plan, and organize, and all the rest. But the main focus of this approach was operational, with these other functions playing a supporting role.

Conversely, workers and low level managers were tasked with contributing to overall strategy by improving the organization’s processes. There was no theoretical line dividing policy from operations; everyone did a bit of both, although the top executives and the line workers were certainly at opposite extremes.

The TWI approach could even be adopted by chief executives of organizations, and in fact it was. Treasury Secretary Snyder used Work Simplification to achieve his policy goal of IRS reform: he whipped up enthusiasm by trusting line workers, worked with them to craft a minimal viable reform, and iterated until he had reorganized and modernized the agency. Empowering the lowest-level employees ultimately empowered the secretary.

For whatever reason, Training Within Industry was abandoned fairly rapidly after WWII. It lasted longer in the federal government, but was abandoned by the mid 1960s.

But it did find a receptive audience elsewhere. Training Within Industry proved popular in Japan; it was adopted by Toyota and rebranded as lean production.

So, a summary: the federal practices of the 1940s and 1950s were based on the immediate forerunner of lean production, a proverbially modern approach to management. This was replaced by a top-heavy approach focused on policy planning; few successful organizations use this approach today.

When reformers suggest abandoning failed reforms, the common response is that we cannot return to the past. The TWI history suggests a response: The past approach is sometimes the more modern and proven approach.

Subscribe now

A previous article of mine discussed the Truman-era reforms to IRS.

One Mr. Andy Brunelle reached out. His grandfather Calvin Wright was nearly the last Senate-confirmed appointee at IRS, being confirmed as the director of their Idaho office in 1951. Following the 1952 reorganization, he had to take a civil service exam to keep his job. After coming second only to a retired customs bureau official, he was able to keep his position at the agency, where he then led a long and illustrious career.

He apparently approved of the 1952 reorganization and travelled throughout the Pacific Northwest to help other field offices successfully navigate it. His work on IRS reorganization continued throughout the 1950s and 1960s, where his speeches and training helped IRS transition to the early computer age.

An episode from 1969 illustrates how long politicization lasted at IRS – Wright’s superiors at Treasury directed him to open his office on a weekend so that a politician could catch up on several years of unpaid income taxes.

Calvin Wright was a sort that you rarely see today: the political civil servant. He spent many years as the auditor of the state of Idaho, where (for the first time in Idaho’s history) he installed modern accounting systems to enable meaningful oversight of their bureaucracy. He unsuccessfully ran for governor; following this electoral defeat, Truman appointed him as director of IRS’s Idaho office.

My prior post discussed the patronage at IRS and said that the Senate intent was “to bolster support [for the income tax] by appointing men of regional renown.” With his extensive experience as an auditor and his statewide reputation, Wright was just the sort of man the Senate wanted (or at least claimed to want).

Unfortunately, there were few others of this caliber, and even fewer who might be nominated via the Senate patronage system. On the other hand, there were plenty of officials who were covetous of expensive fur coats. And so IRS was reorganized, and political appointments there were consigned to history.

He was a man from an era that has doubtless long since passed. The US was better for his service.

The People Respond will be an intermittent series showcasing responses that provide additional context, information, or interesting details. Or perhaps it will be a regular series – that’s up to you.

In any event, entries in this series will be published on Thursdays, whereas regularly-scheduled content will be posted consistently on Mondays.

Jargon and “official-ese” plague governments everywhere. In regulations, reports, and websites, government writing too often avoids the clear and the straightforward. It has always been thus. Consider the following1 quote from the 1890s, both truly painful and quite representative:

Sir: In obedience to your request of the 15th instant for my opinion as to the merit and practicability of the new system of transacting the official business assigned to the office of the Auditor for the Treasury Department, and under my charge, as may have been evidenced to me during the period in which it has been in operation, I beg to submit the following:

In the words of Yes, Minister, “I couldn’t have put it less clearly myself.” Government writing is less stilted today, not least because of the Plain Language movement – an effort to simplify government writing so that ordinary people can make sense of it.

The federal government began2 adopting these principles with an executive order during the Clinton administration, eventually leading to the Plain Language Act of 2010. Today, the major initiative is improving government websites; the US Digital Service, among others, offers resources such as training and style guides.

Bad government writing is an age-old problem, and attempts to improve it are only a bit younger. In 1955, the General Services Administration3 published a writing guide called “Plain Letters”, thereby kicking off an effort to improve writing across the government.

The guide, as the name implies, explained what constitutes a good letter. But more than that, it aimed to train good letter writers. The focus was partially on the final document, but even more so on investing in the writer himself.

Plain Letters

Mona Sheppard – the author of Plain Letters – begins with an anecdote that sets out the problem of bad writing, as well as the remedy:

Once when Franklin K. Lane was Secretary of Interior he happened to see a very ornate letter to an Indian. That letter, in Mr. Lane’s own words, was “so involved and so elaborately braided and beaded and fringed” that he himself could not understand it. So he sent it back to its author with this pith advice: use straightaway English.

We don’t know the rest of the story. But if the author took Mr. Lane’s advice we can easily imagine how he slashed through the trappings that hid the meaning of his letter. He cut out lazy words and cut down big ones. He woke up passive verbs and made them active, straightened out roundabout phrases, and shortened long sentences. Then he tied all his sentences together.

When the author had done slashing and straightening we can imagine that the Indian letter was transformed. From beginning to end its meaning shone clearly in every single word. It may not have been a literary masterpiece, but it was easy to read and easy to understand. It was a plain letter.

Besides the problem and the remedy, these first three paragraphs show how engaging and readable this guide is, at least so far as style guides go. Since there was only a single author, it could have a consistent style and an overall narrative structure – a lesson that government perennially neglects.

But back to the guide: everyone who writes a guide needs to propound some sort of formula, and Sheppard was no exception. She advocates4 a 4-S formula for letter writing: Shortness, Simplicity, Strength, and Sincerity. The first half of the work explains this formula, and the second half gives a series of examples.

The majority of the guide is stylistic: preferring simple over complex words, avoiding the passive voice, and so forth. It is no different from modern style guides in this respect, even if it is better written. However, it was unusually practical compared to modern government style guides.

Plain Letters covers nothing except writing government letters. Being limited to a single narrow subject, it can dive deeper than generic Plain Language style guides. As a result, the guide had two unique strengths: (1) it stressed the need for subject-matter expertise, and (2) it offered advice for improving the substance of letters (in addition to the style).

Knowing the subject

The guide went beyond a focus on good letters, and focused on becoming a good letter writer. That is to say, it focused on investing in the federal workforce; good government writing was evidence that this investment had paid off.

Sheppard advises government employees to master their agency’s business, and (if they haven’t) suggests practical ways to go about learning.

She begins5 with the need for subject-matter expertise and offers a concrete example of how it can lead to clearer writing:

Know your subject so well you can discuss it confidently and naturally. […]

A government letter may turn out hard to understand when the writer, not quite sure of his subject, plays safe by sticking to the language of rules and laws.

Let Letterwriter X show you what happens. Not sure of the meaning of a regulation he wrote this sentence in a memorandum:

“PA” means that you are classified with those employees currently serving under absolute or probational appointments in positions held by the employee on a permanent basis, including preference eligibles in excepted positions under appointments without time limitations.

Mr. X’s boss, sure of the meaning of “PA” and the status of the person to get the memorandum, might rewrite that sentence to say:

You are on the “PA” list because you have a permanent civil service appointment.

And how should a bureaucrat acquire this knowledge? As part of a longer list of suggestions, she offers6 three suggestions to civil servants about acquiring subject-matter expertise:

1. Study all new laws and rules that you will be called on to explain. If you are not sure of their meaning, ask somebody who is. And don’t wait until you get a letter on the subject. Be ready.

2. Discuss new subjects with your fellow employees.

3. Learn the practical art of clearing up knotty problems with examples. Do this by making up questions and answers in advance. You’ll find the ready-made examples handy in writing letters.

Once employees had acquired subject-matter expertise, they could put it to use and write letters with better content.

Better substance

Plain Language guides today often focus on using unambiguous language in e.g. regulations. In other words, they generally focus on adding clarity to given ideas. Since Sheppard is dealing with a narrowly-defined issue, she can suggest ways to improve the substance of letters, thereby avoiding common mistaken approaches.

For example, she suggests7 a thorough rewrite to a stilted and cold opening paragraph. It adds entirely different (and more helpful) content, which draws upon the subject-matter expertise of the letter writer:

Before:

This will acknowledge receipt of your letter of September 5, 1954, enclosing an application for --- and stating that this application was previously submitted to this agency on August 25, 1954, but was returned without evidence of any action having been taken thereon.

After:

You are due an apology for our error in returning your --- application as we recently did. You had every right to expect a letter of explanation. The fact is, your application had to go back to you for more information, but somehow it got into the mails without our letter telling you this.

The guide covers several issues that are specific to government letters: crafting a strong beginning, or acknowledging mistakes. One8 piece of advice: “Admit mistakes. Don’t hide them behind meaningless words.”

And, like most style guides, it gives a series of examples – good letters, generally by famous officials past and present.

One was FTC Commissioner Lowell Mason’s letter to a concerned citizen. Mason’s response9 was so charming that I will indulge in one last block quote:

Thank you for your letter of June 18 stating the reasons for maintaining the present status of Section 2 (c) of the Robinson-Patman Act.

They are the usual arguments in favor of this law, but you have expressed them so well, so carefully and in such an open and friendly manner that I confess not only admiration for your presentation, but complete agreement with many of your points.

There is, however, more to this problem than either you or I have covered. With your leave, I shall search out in a later note some of the pros and cons of this most interesting subject we might consider together.

June has been a heavy month for me, so I expect to take a short vacation, what Walt Whitman described as the “white spaces in life.” You will hear from me the latter part of July.

Plain Lessons

Plain Letters is, in the end, just a style guide. It is an unusually well-written one, but much of the content is no different than what we have today.

And in most respects the modern approach is better. The biggest problem with Plain Letters is its overly literary style, which uses too many complex sentence constructions. While it is an unfortunate fact that the average adult reads at an 8th grade level, it is nonetheless a fact, and modern Plain Language guides rightfully take account of this.

But it holds enduring value as part of the 1950s investment in the federal workforce. This guide was10 distributed to hundreds of thousands of civil servants, thousands of whom attended 4-S training sessions.

This focus was reflected in the guide itself. It doesn’t just focus on better letters, it focuses on better letter writers. It tells government employees squarely that they have got to know what they’re talking about, and then gives them some pointers on how to learn. Writing guides today could focus on the person writing documents, and not just on the documents themselves.

Government writing guides today could also dive deeper into specific issues: they could offer advice on substance, not just style. Plain Language guides on, say, regulation might be written by regulatory commissioners. They might offer advice about specific phrases that lead to litigation, and better ways of rephrasing them.

Finally, Plain Letters showed that guides with a narrow scope and a single author can be readable and memorable, which is strikingly unlike style guides generally.

Perhaps some enterprising US Digital Service employee will write Plain Emails.

Appendix (source): Do you want to write gooder, too?

Source: The book is freely available on Google Books. I recommend it highly: it has been helpful for my long-form writing and has dramatically improved my emails. There’s a good bit more to the guide than I have described.

Before you go

A question for my international readers: is “government style” a worldwide problem? Does the “before” quote in Better Substance remind you of bureaucrats in your country, with its overly-formal, cold, and dense language? If not, what is their “government style”?

And a request for my readers: Does anyone know anything about Mona Sheppard?

There is a bit of biographical information available about her, and apparently she wrote another book on business letters.

Over the past few weeks, several readers have reached out to offer additional information or context about previous posts (and more on that soon).

So, here I will preemptively ask.

State Capacitance launched two months ago. Since then, I’ve written some articles and gained some readers. You are, presumably, one of them – thank you! For my part, what have I written so far? And what comes next?

2024 in review

The first pieces were deep dives into the history of government reorganization. In particular, I told a tale of two reorganizations: the midcentury reorganizations of the US Department of Agriculture (USDA), and of the Treasury Department.

Each reorganization began with an idea – reformers had some theory about how bureaucracies should be organized, which they then applied to reorganize an executive department. Accordingly, the story of each reorganization is told in two parts: the idea behind the reforms, and then the reorganization itself. What were these pairs of articles?

• First, a bad idea – functional reorganization. This principle held that agencies ought to correspond to functions (like research and regulation), rather than to subjects (like entomology, or forests). The USDA reorganizations of the 30s-50s applied this idea dogmatically, with minimal planning or feedback, and indeed with a poor conception of what problem they were trying to solve.
  
  The reorganization cleaned up the department’s org charts, but at the cost of the department’s independence and competence. After the rationalistic reorganization, the farm lobby had wrested control over agricultural policymaking from the previously quite independent department.

• Second, a good idea – work simplification. This method was developed in WWII, and held that managers ought to be trained to simplify department procedure and cut red tape. The Treasury Department reorganizations of the 1940s applied work simplification in a series of steps: they began with small teams that made minor improvements, and built up to major departmental reorganization.
  
  Through applying work simplification, the IRS brought its work up-to-date, cracked down on corruption, and modernized its archaic IT. Eventually, Treasury used these successes to build up support for abolishing Congressional patronage in the department.

These case studies are more than interesting history – effective approaches like work simplification can still be adopted today. Further, case studies of both successful and failed reorganizations help us to evaluate contemporary proposals for reforming the government.

The publication also featured a historical article on how cities and states spent money before they adopted budgeting.

What next?

The publication will continue featuring case studies: deep dives into how agencies were founded, or reformed, or made worse. There will also be case studies on specific processes – for example, what did agencies do to control paperwork before the Paperwork Reduction Act?

The publication will feature more historical articles. In particular, this year will feature an extended series of posts on the Progressives and the Progressive-era reforms. Their ideas were foundational to the modern American bureaucracy, but the discussion of their specific reforms is often superficial and inaccurate. And it’s worth knowing some of their failed proposals, too – many were unique ideas for dealing with issues we still face.

There will also be a new type of content: notes on public administration. For example, there are many proposals for (say) changing the budget process. How can we classify different approaches to budgeting in order to make sense of the many proposed reforms? The notes on public administration will provide the necessary concepts for making sense of all these proposals.

And a final type of content for the new year: whatever you’re interested in. If you could know anything about the history of government administration, what would you want to know?

In the scandal-plagued Truman administration, the most corrupt agency was the Bureau of Internal Revenue (now IRS). After a series of lurid scandals, involving officials being bribed with luxuries such as fur coats, the Bureau of Internal Revenue’s commissioner was sentenced to five years in prison. More than 150 other bureau officials were fired.

Nor was the public much impressed by the routine work of the bureau: corporate tax investigations were two years in arrears, while income tax refunds took more than twelve months to arrive. This agency, with its 19th century organization and procedure, was obviously not up to the job in postwar America.

Congress was inclined to pile on new procedural controls in response to the corruption. But Truman’s Treasury Secretary, John Snyder, saw that corruption and inefficiency were symptoms of the same problem: bad management. Better management would allow the agency to root out corruption while also modernizing its archaic business methods. To strengthen management, the Bureau needed fewer procedural controls, not more.

If Snyder wanted to convince Congress to go along with this, he needed to demonstrate that trusting the agency could pay off. The Treasury, therefore, needed put its own house in order. It turned to work simplification.

My previous post discussed work simplification: the federal government’s approach to management, which they developed during WWII. Although this approach was invented during the war, the Truman administration was the first to apply it to an entire department. Treasury Secretary John Snyder aimed to straighten out his department through, as a guide to work simplification put it, taking advantage of “the great reservoir of unused practical knowledge” of frontline managers.

Within five years, the Bureau had brought its work up to date and cracked down on corruption. They had been so successful at adopting better management that the once-decrepit Bureau had become a world leader in adopting computers! Congress changed its mind about new procedural controls and instead accepted the administration’s wholesale reorganization proposal, which finished the job of abolishing corruption in the Bureau.

The Growing Pains of Taxation

Before WWII only a small number of people paid income tax. During the war, demands for revenue resulted in a vast expansion of eligible taxpayers, with the number of tax returns rising from 20 million to 80 million, and the amount of money collected rising from $5 billion to $40 billion.

What sort of an organization did the Bureau of Internal Revenue have to confront this task? Most work was managed by 64 regional “collectors” of revenue, who (rather surprisingly) were Congressionally-confirmed appointees. Ever since the creation of the income tax, people who filed income tax made payments to a patronage hire! The original intention during the Civil War was to bolster support by appointing men of regional renown; regardless of if this scheme ever had any merit, it was wholly unsuitable for the 20th century and almost inevitably led to corruption.

The bureau was thus riddled with a long-established system of patronage that made effective management impossible, as the collectors were more loyal to their Congressional sponsors than to their ostensible bosses.

The rest of its organization was not much better. The agency was generally organized according to the tax collected, so that each type of tax – such as income, excise, gift, etc. – had its own field offices and reported separately to the Bureau’s commissioner. Not only was this inefficient for management, but taxpayers had to pay different taxes to different offices, and perhaps had to visit multiple far-flung offices.

These problems had always existed but were exacerbated by the war. And the war brought more problems than just an expanded workload: the Bureau lost most of its trained workforce to the war effort, with e.g. accountants being drafted into the military or hired by war plants. With this reduced manpower, paying taxes was essentially voluntary – the Bureau couldn’t investigate more than a tiny fraction of returns, and in fact couldn’t even process returns on time. Nor could they investigate charges of corruption against their (frequently unqualified) workforce.

By the end of the war, work was hopelessly in arrears while corruption was an increasingly open secret. John Snyder – Truman’s Treasury Secretary – was determined to set things right.

Incremental Improvements

When Snyder took office in 1946, he saw managerial reform as key to fixing both corruption and inefficiency. Convoluted procedure prevented good employees from accomplishing anything, while also shielding corrupt officials behind the process’s obscurity. He immediately began studying ways to improve the Bureau of Internal Revenue’s management.

He gathered the top officials and tasked them with improving the Bureau. More than that, he wanted to involve the rank-and-file immediately, stating:

In addition, two other important programs were put into operation in 1947: One was a work simplification program designed to take advantage of the operational know-how of the lower level Bureau employees. The other was a cash awards program in which we rewarded employees who made time and effort-saving suggestions. By the end of 1951, five years after the programs were initiated, more than two thousand cash awards had been made, and about twenty-two hundred work simplification program suggestions had been put into effect.

The department made $38,000 available as rewards for good suggestions. Or after adjusting for inflation, $500,000 – far from petty cash! But the key initiative was a Work Simplification program: front-line managers were trained to trace their office procedure from beginning to end, making simplifications at every point (as my previous post describes in detail).

The Treasury’s Work Simplification initiative began in 1947, with a pilot project in St. Pauls, Minnesota. Treasury got feedback on its training, and gradually expanded the program throughout the Bureau and eventually the entire Treasury department. Through the feedback of low-level managers, the Bureau of Internal Revenue made thousands of improvements to streamline procedure.

Most were minor improvements for handling internal office paperwork. Some were straightforward, yet highly visible to taxpayers: the Bureau combined related forms such as Social Security payments and income tax withholding. It began tracking income tax refunds, to bring the wait down (from more than a year) to a month. And some of these improvements were daring: in 1948, for example, a Cleveland office became the first office to experiment with punch-card equipment.

This is not to say that the political leadership was simply deferring to the wishes of career civil servants. Snyder knew from the beginning what he wanted to accomplish. But he saw the merits of beginning with narrow goals and winning fast, achievable victories:

None of the reforms so instituted were designed as cure-alls. None tried to create a tax man's utopia. They were day-to-day, realistic improvements that could be carried out within the Bureau without congressional approval, and without upsetting revenue collections, and they were carried out within budgets approved by the Congress.

After the system of Work Simplification with cash rewards was well underway, the Bureau employed a consulting agency to study the Bureau’s management. Snyder felt that doing things in this order was crucial: External consultants should be employed, but only after interest in reform had been stirred up among the rank-and-file. The consultants should provide cover to enact more radical reform, but they shouldn’t be a wholly external force that harassed the career employees.

Crisis and Radical Improvements

These consultants were hired in late 1948 and issued their report in 1949, mainly suggesting ways to adopt then-modern technology and to further streamline office procedures. But after this inoffensive report, the Bureau soon received a far more bitter and searching investigation at the hands of Congress.

In 1950 and 1951, Congressional oversight uncovered the depth of corruption at the Bureau, resulting in charges being filed against several officials and many more being fired. Snyder himself was certainly not blameless: he had been very slow in firing political appointees with credible accusations against them.

Crucially, however, Snyder could accurately say that he had already recognized the problem and had been reforming the Bureau for several years. More than that, he could point out that worst corruption was caused by Congress, who created (and benefited from) the Bureau’s patronage appointments. This was, in Congress’ view, an uncomfortably good point.

Because Snyder had already achieved significant successes, he was able to channel the Congressional outrage in a productive direction, later stating:

The investigations that followed had a paradoxical effect. They were to prove in the long run a political liability to the Truman administration, but they were also to make more readily acceptable the wholesale reorganization of the Internal Revenue Service which I so greatly desired. I was convinced that without the investigation, there would have been violent opposition and small chance of approval in Congress of any plans to abolish political appointments in the Bureau. The investigations undoubtedly made many Congressmen reluctant to oppose the reorganization.

In 1952, Truman proposed Reorganization Plan Number 1, which abolished the patronage position of revenue “collector” and reorganized the Bureau generally. Congress accepted this reorganization plan1, allowing the Bureau to implement the far-reaching plan for reform that they had developed (described below).

A Scheme for Reform

As the Congressional investigations began in 1950, the Bureau continued its plans for reform. The consultants had submitted their evaluation to the Bureau, which had to decide what to do with it. They decided that the consultant’s report would be – another crucial point – purely advisory. It was taken only as a document for discussion, with the Treasury department being free to accept or reject any particular point. Eventually they hashed out the following plan for reorganization:

First, under the new plan they created a long-overdue inspection service and began systematically training field officials. They also, as stated, abolished the patronage appointments. This dealt with the immediate scandal and went a long way toward ending corruption. But to enable better management generally, they fundamentally reworked the Bureau’s organization.

Previously, each type of tax had its own set of field offices. Under the new plan, they grouped everything by function2: all appeals were handled by one division, all enforcement by another, and all field offices were consolidated so that taxpayers only had to deal with one regional office.

However, the alcohol and tobacco taxes were exempted from this general rule, as their purpose was more akin to suppressing vice rather than collecting revenue per se. Agency officials noted3 that their intent in reorganization was “Not to make a fetish of rigid consistency in the face of inconsistent circumstances.”

After long discussion, they felt that their plan was ready for adoption. The Congressional investigations gave them their chance, and this was the plan that Truman enacted in 1952.

Implementation

They Bureau had already begun implementing what (relatively little) reorganization it could accomplish without statutory changes. The Congressional investigation gave them the chance to propose the necessary statutory changes, which was accomplished via Reorganization Plan No. 1 of 1952.

The leadership felt, however, that changing the statute was only the beginning. Congress had abolished patronage, but Congress might bring it back. If the leadership wanted this crackdown on corruption to stick, they needed to prove that the reorganization was successful. That is to say, they still needed to deal with the problem of implementation.

To prepare, agency leadership formed small committees, who in turn prepared instructions on transitioning to the new system. They prepared a playbook covering everything from new methods of accounting, to new personnel practices, to disposing of old and outdated forms.

Even with these elaborate preparations, agency leadership decided to implement their reforms in phases. First, they decided to try a pilot program at a single (notoriously corrupt) field office in Chicago. The most senior Treasury officials, including the Secretary himself, addressed the Chicago office on the day of the transition and promised close support for several weeks.

As leadership learned what worked and what didn’t, they updated their plans and their written instructions accordingly. They also discovered the need for things such as legal documents conferring authority on the new officials. These were accordingly added to the implementation guides so that future transitions would go more smoothly.

The Bureau scaled this process up to include more field offices throughout 1952. As their transition plans became better tested, their playbook was refined; eventually, transitioning offices barely needed any support besides these written guides. After leadership felt they had a handle on the problem of reorganization, they reorganized the Washington headquarters itself.

In addition to reshuffling the org chart, earlier experiments with technology became mature enough that reformers could take advantage of them. The Cleveland office had pioneered the use of punch-card machines; this trial was scaled up to include more offices and more forms. Similarly, photocopiers and mimeographs were rapidly adopted across the Bureau, while microfilm was used to save4 “literally acres of storage space”. They even began adopting the primitive computers that existed in the early 1950s.

In 1945, the Bureau’s procedure had barely changed since its founding5 in the Civil War. By the 1952 reorganization, the Bureau was a world leader in the adoption of computers.

This reorganization happened in the waning days of the Truman administration. The outgoing administration looked with pride on its accomplishments, but had a sober view of the remaining difficulty, stating6 “The work of reorganization will not be completed even after all of the new districts have been established, however, for the task of implementation in each district still will be continuing.”

What Snyder Can Teach Us

What can we learn from all of this? First, some lessons about policy implementation.

Snyder had very ambitious goals from day one: adopting modern business methods and banishing corruption. He saw these as two sides of the same issue, that of improving management. The Bureau’s unwieldy organization, with its overelaborate and illogical procedures, resulted in delays while also shielding corrupt officials from accountability. By contrast, clear organization and streamlined procedure would result in fast service with effective oversight. His ultimate goal was to abolish the Congressional patronage in the Bureau that gave rise to so many of these problems.

But he chose to start with the simplest goals, rather than with grand schemes of reform. He built up enthusiasm among line managers through having them win fast victories by improving existing processes. This laid a foundation of success that won him the confidence needed to carry out larger reforms.

The Bureau’s success with these larger reforms was based on a tight feedback loop between their planners and their implementing teams. Leaders followed the first implementations very closely until they felt confident that their playbook could scale. At every point they solicited feedback, so that each new implementation was more successful than the last. They saw that formulating policy was inseparable from implementing it: their reforms had to publicly succeed in order for them to last.

And thanks to their efforts, their reforms did last. Congress has never restored patronage at the IRS; more than that, nobody has even seriously proposed it. While IRS has had scandals since the 1940s, it has never had that particular sort of corruption again, which is as much success as one can reasonably ask for.

Second, some reasons for optimism.

Truman, Snyder, and the 81st Congress were by no means perfect reformers, the sort whose equals we will never encounter today. Truman and Snyder were both far too trusting of corrupt friends, while Congress was mainly focused on scoring cheap political points in response to bad headlines. Yet with the right approach they managed to do the right thing.

It also shows how quickly an agency can move from stagnation to the frontier of technology. The Bureau had modernized almost nothing since its creation, and then adopted the best of IT over the course of five years. The IRS today is attempting to update its ancient codebase; it could do much worse than to learn from Snyder.

Replacing COBOL is, of course, not quite the same as adopting mimeographs. But the principles of incremental, agile improvement are always and everywhere the same.

Appendix: Sources

Two Treasury officials wrote contemporaneous articles explaining how these reforms were accomplished: Treasury Secretary Snyder himself, and Assistant Secretary Parsons. Facts about the reorganization’s timeline or about specific actions are drawn from them. I cite quotations.

Sources:

1. Snyder, John W. “The Reorganization of the Bureau of Internal Revenue.” Public Administration Review 12, no. 4 (1952): 221. https://doi.org/10.2307/972486.

2. Parsons, William W. “Installing Management Improvement in the United States Treasury Department.” Public Administration Review 10, no. 3 (1950): 176. https://doi.org/10.2307/972813.

Snyder much later gave interviews that covered the reforms. Assertions about Snyder’s motivation, thinking, or purpose are generally drawn from them. (But with caution, since he was minimizing his personal responsibility for the scandal.) I cite quotations via links.

Source:

1. “Jerry N. Hess, ‘John W. Snyder’, Harry S Truman Library”.
   
   The relevant interviews are 21 through 23.

As this is only a blog, I have exercised my right to streamline procedure in providing citations.

For the federal bureaucracy, the 1940s through the 1960s are a nostalgic time. The era saw one spectacular achievement after another: from winning World War II, to building the interstate highway system, to landing on the moon. At its high point, trust in the federal government reached almost 80% in the 1950’s, as opposed to only 20% today.

Trust in the federal government has plummeted alongside the federal government’s ability to accomplish anything – which is no coincidence. Although government competence has changed for many reasons, there is one forgotten reason: after the second World War, the government was competent because it taught its managers to be competent.

Subscribe now

During World War II, the poor management in the federal government was keenly felt. Although federal management had never been especially good, it reached a boiling point when it began noticeably impeding the war effort. The Bureau of the Budget (now OMB) responded by creating a new management unit tasked with training federal managers.

They termed their newly-developed management approach work simplification, which held that implementation and policy went hand-in-hand, and therefore managers had to be trained to streamline procedure in order to achieve policy goals. Moreover, the Bureau of the Budget felt that this viewpoint could be systematically taught to federal managers of average competence, and developed a training program to do so.

Work Simplification

During the war, the civilian agencies were incredibly short staffed due to the draft, so any procedural red tape or poor distribution of work created instant bottlenecks. Many of these bottlenecks directly impacted the war effort, as (for example) with slow approvals for important construction projects. The Bureau of the Budget therefore began an initiative to improve management around 1942.

They conducted user research with several agencies and eventually felt they had a management system that could scale, which they termed Work Simplification. They taught managers Work Simplification at training seminars, and also created guides and pamphlets to distribute across the government. I quote from one of their guides1 that sets out the problem, the audience, and their goal:

Thinking of this sort has been going on in the United States Bureau of the Budget for some time. It has culminated in the decision to make a concerted drive to capture the best available means for exposing and disposing of common management problems, set it forth in clear, simple language, and put it in the hands of those who can use it to best advantage. And who are they? They are the operating managers of government: middle management people and first line supervisors. […]

From the standpoint of the Bureau of the Budget, Work Simplification is a method of attacking the procedural problems of large organizations by equipping first line supervisors with the skill to analyze and improve procedures. It provides a way of tapping the great reservoir of unused practical knowledge represented by this group.

Their management agenda developed a training program for the managers closest to the ground, rather than (as is common today) focusing on top leadership. The guide elaborates on their approach and its value:

In this program, operating agency people who have been instructed in the plan train agency first-line supervisors to study and solve the basic problems of their own units. Therefore, improvements grow from the “grass roots”, and management obtains results which cannot be achieved in any other way. Supervisors are taught to gather relevant facts quickly, to organize them in simple chart form, and to interpret them properly. […]

However, the cycle is not complete when the supervisors have merely mastered the use of the techniques. Every supervisor in each training group should successfully apply the techniques to the work of his office and come up with an improvement which is good enough to be adopted and installed. This is an extremely important feature of the program. It immediately produces a dividend—a saving in time, money, and manpower—thereby earning continuing support from the top management of the agency concerned. Equally important, with a success story on his record each supervisor is stimulated to continue to apply these techniques to his work problems. In the long run, it is this continued application of the analytical approach which will produce the really big dividends in improved federal management.

There were several elements to Work Simplification, more than could be discussed in a single post. This post drills down into a single and quite narrow issue, namely how managers were trained in process charting to simplify the flow of paperwork through an agency.

Process charting

At a Work Simplification training seminar, managers would be taught to create several types of charts for analysis, one of which was the process chart. The chart would show the flow of an application (or other document) through an agency from start to finish. But the purpose was not understanding procedure for procedure’s sake. To continue quoting the guide:

The process chart is a device for tracing and highlighting work flow. To make such a chart it is first necessary to identify an office procedure involving a number of steps in sequence. Usually a form, a paper, a case or other office medium is selected to be followed through the several steps in processing it. In some cases it may be the steps performed by a single individual that are recorded. In either event the steps are recorded in order on a special form which is provided for the purpose. […]

THE PROCESS CHART shows you the “who”, “how”, and “when” of a whole work process and permits you to ask “why” about every step. And only by asking what is the purpose of every step can you find ways of simplifying procedure, getting rid of bottlenecks in your unit and smoothing out rough spots.

This approach addressed one of the most common frustrations with manager’s approach to paperwork: federal executives are routinely experts in how a process is done, but equally often neither they – nor anyone else – knows why each step is done. The training forced managers to think about the purpose of individual steps, and consider if the process as a whole achieved the desired outcome.

To be concrete, a process chart traced the lifecycle of a document and classified each step of its lifecycle into the following categories: (1) being created/changed, (2) being transported, (3) being stored awaiting further action, or (4) being checked/verified. Each of these categories was denoted by a particular (frankly unintuitive) symbol. A manager would transcribe the step-by-step lifecycle of a document and label each step with the appropriate symbol.

An example will likely make this clearer. The selection from the guide below shows both an example process chart and the sort of conclusion they might want managers to draw:

The purpose of understanding the process was to simplify the process. In particular, a government process often has individual safeguards that seem sensible in isolation, but that taken together waste a great deal of time in return for minimal gain. Managers were encouraged to consider the process as a whole and drop controls that failed to add value. Reformers today might well adopt as a motto “Don’t pile controls too high; take a calculated risk. It may pay!”

Lessons today

Although Work Simplification was developed during World War II, it was still the common approach for training federal managers into the 1960s. These were the stodgy managers of the Eisenhower era who oversaw the building of the interstate highway system, or the administration of the GI bill.

This is not how the federal government approaches management today. It would be, obviously, unreasonable to claim that earlier success was entirely due to training managers differently. But it clearly contributed – their methods explicitly aimed to solve issues that today’s processes aggravate.

In particular, the Bureau of the Budget’s work almost remarkably anticipated current conversations on government efficiency. Reformers note that the bureaucracy piles up layers of procedure without ever rethinking them – process charting taught managers to reduce procedural burden. Reformers note that government IT piles up layers of software from different eras, with nobody understanding how it fits together – process charting taught managers the start-to-finish viewpoint. Reformers note that bureaucrats rarely consider what it’s like to actually apply for benefits – once again, a failure that process charting aimed to correct.

Process charting is clearly not a perfect solution to any of these issues. But it is proof that the government can train bureaucrats to tackle these issues head-on!

The overall lessons of Work Simplification are even more important. Work Simplification’s success did not last forever, but it did last for several decades. And it achieved its success because the Bureau of the Budget created free training for low-level managers, while nobody else particularly cared.

So would-be bureaucratic streamliners today – proponents of product management thinking, agile IT development, or what have you – might imitate Eisenhower’s bureaucrats. Above all, they should prove that their proposals are a rational method that can be systematically taught to low-level managers, in order to put their “great reservoir of unused practical knowledge” to use.

Appendix: Do you want to simplify government, too?

This piece discusses only one part of the government’s training initiative (although it was the most important part). The training manual is called “Work Simplification as Exemplified by the Work Simplification Program of the U. S. Bureau of the Budget.”

The full manual is excellent, and well worth reading. It is online on Hathitrust. To read it, you have to be located in the US. To download it, you have to have an institutional subscription.

Alternative, I can send you a copy – it’s in the public domain. If you’d like a copy, message me on Substack, LinkedIn, email (khawickhorst@gmail.com), whatever. Just include your email address and the fact that you’d like the book.