The Politics of Corruption

Lecture 6: Patronage and the Progressive Era in the US

Bogdan G. Popescu

Tecnológico de Monterrey

What just happened?

Diego worked three months of evenings and was paid nothing for it
The party won, and a public job came free
No money changed hands. Nothing was hidden. Nobody was embarrassed

He paid in evenings. The party paid with a public job.

This arrangement has a name: patronage — public jobs handed out as rewards for political support, rather than for being able to do the work.

Is that corruption?

Giving out public contracts

The same office, quiet. The official hands a signed City of Northland contract agreement across the desk to a contractor carrying rolled building plans. No money is visible.

The contractor holds up a city invoice for $1,000,000 stamped APPROVED, beside a sheet showing the true cost of the project as $800,000.

The official hands over the signed contract while taking a thick bundle of banknotes beside a party collection box.

A private company is tasked to renovate a public hospital.

Is this corruption?

The work costs $800,000. The city is billed $1,000,000.

The city pays the extra $200,000 — out of public money.

The contractor is the mayor’s brother — and part of the $200,000 goes back to the party.

Nobody asked a citizen for a bribe. The public paid for both.

The Puzzle — and Today’s Route

A job handed out for campaign work. A hospital billed $200,000 too much. One of them is plainly a crime. Suppose the other is perfectly legal.

Would that make it not corruption? That is the puzzle — not which one is illegal. Nobody in that office had to be unusually greedy, and both favours came out of the same desk.

Four questions:

Is any of that corruption — and is it all the same kind?
If the favour is the currency, how do you make it worthless?
But a city still has to hire someone. How do you stop the job being sold?
Did any of it change what governments actually did?

Corruption is not spread evenly

The United States is relatively clean by global standards today — although far from corruption-free.

The United States did not start clean

New York City, 1871: the same arrangement, at city scale
Even the police were in on it — the prosecution had to come from the state

Not one dishonest official. A system — so removing one man would not have fixed it.

Can a country substantially reduce corruption once it is embedded in its political system? The United States is one case where we can watch.

Thomas Nast's caricature of Boss Tweed: a heavy man in a suit with a money bag marked with a dollar sign in place of his head.

The machine’s boss, William “Boss” Tweed, drawn by Thomas Nast, Harper’s Weekly, 1871. On prosecution coming from the state rather than the city, see Glaeser & Goldin (2006, p. 12). Library of Congress.

Who was Boss Tweed?

Photograph of William M. Tweed, a heavy bearded man in a dark three-piece suit with a watch chain, hands in his pockets.

William M. Tweed, 1823–1878, New York
He was never mayor. He chaired Tammany Hall, the city’s Democratic party organisation
His power came from running the party, not from holding high office
Tammany gave poor and immigrant New Yorkers jobs, coal and help with paperwork — in exchange for their votes
And it took a share of every city contract

So which kind was it?

The inflated bills were venal — private profit out of public money.

The machine itself was systematic — jobs and contracts allocated to keep Tammany in office.

Both, in one organisation.

Tweed was convicted in 1873 and died in jail in 1878. Tammany went on running New York for another sixty years.

Biography is background, not from the assigned readings. On prosecution coming from the state rather than the city, see Glaeser & Goldin (2006, p. 12).

What Nast actually drew

Nast's caricature alone: a money bag with a dollar sign in place of the head, a diamond pin at the throat.

The photograph of Tweed beside Nast's caricature of him, same stance, same build.

Thomas Nast, Harper’s Weekly, 1871
Tweed’s head is replaced by a bag of money
The joke is not that he was greedy. It is that the “brains” behind the machine was not cleverness — it was cash

Nast ran these for months. Many of the machine’s voters could not read English. All of them could read this.

Though the exposure was set off by a rival politician, not by the press alone (Glaeser & Goldin, 2006, p. 12).

And this is why he is the famous one. Other cities had machines just as durable. They did not have Nast.

Two ways of drawing the same man

Three figures side by side: the photograph of Tweed, Nast's money-bag caricature, and the party boss from Diego's ward office in a blue suit.

The photograph. The cartoon. And the boss from Diego’s ward office.

Nobody in the first picture looks like a criminal. That is exactly why the machine worked — it looked ordinary from the inside, and it took a cartoonist to draw what it was.

Part I — Two Kinds of Corruption

Wallis: the word we use today is not the word they used.

Two kinds of corruption

Which way is the arrow pointing?

%%{init: {"theme": "base", "width": 1200, "height": 440, "fontSize": 16, "themeVariables": {"fontSize": "16px", "fontFamily": "Helvetica, Arial, sans-serif", "primaryColor": "#eef2f6", "primaryTextColor": "#1e293b", "primaryBorderColor": "#334155", "lineColor": "#b44527", "edgeLabelBackground": "#f9fafb", "tertiaryColor": "#ffffff"}, "flowchart": {"useMaxWidth": true, "htmlLabels": false, "nodeSpacing": 90, "rankSpacing": 220, "padding": 28, "curve": "basis"}} }%%
flowchart LR
  A["PUBLIC OFFICE<br/>the power to decide"]
  B["ECONOMIC BENEFIT<br/>money, jobs, contracts"]
  A -- "VENAL<br/>office sold for private gain" --> B
  B -- "SYSTEMATIC<br/>benefits buy political power" --> A
  classDef box fill:#eef2f6,stroke:#334155,stroke-width:2px,color:#1e293b;
  class A,B box;

The contractor’s kickback was the first. The party’s control of the jobs was the second. Both in that one office.

But they do not cost the same.

Venal corruption moves money between people.
Systematic corruption keeps everyone outside the coalition out of the market —

“Systematic” is his term; it does not mean “widespread” (Wallis, 2006, pp. 25–26).

So is this a new box, or a new question?

Neither replaces what we already have. Each framework asks a different question about the same office:

The question	Where it came from	Diego and the contractor
When can an official do this?	monopoly + discretion − accountability	one party controls the jobs and the contracts
Which way does the exchange run?	venal / systematic (today)	both directions, at the same desk
Why does everyone go along?	norms and expectations	nobody hides it, nobody is embarrassed
Who could stop it?	principal–agent vs collective action	even the police were part of it
What does it cost?	grease or sand	the public pays the extra $200,000

So venal/systematic is not a rival label. It answers the question the others leave open: what is the corrupt exchange for?

Klitgaard tells you when an official can. Wallis tells you what he is doing with it — lining his pocket, or keeping his party in power.

Exercise 1 — Which kind?

In pairs. Venal, systematic, or both — and say why.

A customs inspector waives an import duty for a cash payment.
Every firm that wins city construction work also donates to the governing party. Nothing is hidden — and suppose it is entirely lawful where this happens.
A mayor buys land along a subway route before the route is announced.
Only banks whose directors back the governing faction receive charters.

Then two questions. Which of these are still a problem the morning after you jail everyone involved and replace them with honest people?

And suppose number 2 is perfectly legal. Would that make it the least worrying one here — or the most?

Exercise 1 — one question sorts them

What is the exchange buying?

Case	What does it buy?	Kind
1. The inspector takes the cash	a waived duty, and his own income	Venal
2. Winning firms donate to the party	the party’s hold on office	Systematic
3. The mayor buys the land first	a private profit	Venal
4. Charters only for the faction’s allies	a coalition that keeps the faction in power	Systematic

Not “who got richer”. Systematic arrangements make individuals rich too — Tweed died wealthy. The question is what the exchange is for.

Now jail everyone and put honest people in every chair. 1 and 3 need somebody dishonest, so they stop. 2 and 4 carry on — the arrangement does the work, not the character.

And if number 2 is lawful, that makes it harder to deal with, not safer: there is nobody to prosecute, and the market stays shut to everyone outside the circle.

Part II — Taking Away What There Is to Sell

Wallis: an economic solution to a political problem.

Imagine you need a politician’s permission

Imagine you want to open a bank.

Today, you would normally meet a set of rules and apply.

But in the early United States, that was not enough.

You needed the state legislature to give you permission

Politicians could decide who got permission and who did not

That made permission extremely valuable

People who received the favour had a reason to support the politicians who gave it

Nobody needs to take a suitcase of cash. The permission itself is the favour.

Why is that a corruption problem?

Imagine there are ten people who want to open banks.

But politicians choose only two.

Those two receive something valuable
The other eight are kept out
The winners now have a reason to support the politicians
The politicians have a reason to keep deciding who gets in

The system can sustain itself even if nobody personally takes a bribe.

That is Wallis’s idea of systematic corruption.

Then the system ran into a crisis

In the 1830s, many state governments borrowed heavily to finance banks, canals, railroads, and other projects.

Governments suddenly owed enormous amounts of money
By 1842, eight states and Florida had stopped paying some of their debts
Legislatures investigated what had gone wrong
Theft explained some cases — but not most of them

The bigger problem seemed to be the rules themselves:

politicians had too much power to decide who received valuable economic privileges.

The crisis did not automatically create reform. It made the old system much harder to defend.

The move: remove the bargaining chip

Before

You want to start a company → politicians decide whether you may do it.
Permission is scarce, so political connections matter.

After — general incorporation

You want to start a company → meet the published rules and you may do it.
Politicians no longer choose you personally.

The reform did not prohibit every corrupt bargain.
It removed something politicians could bargain with.

Wallis (2006, p. 50), in Glaeser & Goldin, eds., Corruption and Reform.

So what?

Systematic corruption needs something valuable that politicians can selectively give away
General incorporation attacked the system by removing politicians’ discretion over entry
The lesson was not “punish corrupt politicians” — it was make the favour impossible to sell
But politicians still controlled other scarce benefits
Most importantly: government jobs

Take away one political currency, and the question becomes:

what else can the party use to buy loyalty?

Part III — Machines, Spoils, and the Long Fight for Merit

Patronage, and the long argument about how to end it — the Progressive Era, roughly the 1890s to the 1920s.

Tammany Hall, 14th Street, New York, 1914. Photograph by Irving Underhill. Library of Congress (LC-USZ62-101734).

Back to Diego

The ward office. The official hands Diego his appointment letter: you worked the campaign, the water department needs a clerk.

No money changes hands. Nothing is hidden. Nobody is embarrassed.

So why would the party give Diego the job at all?

Because the job is the fuel

%%{init: {"theme": "base", "width": 1150, "height": 560, "fontSize": 16, "themeVariables": {"fontSize": "16px", "fontFamily": "Helvetica, Arial, sans-serif", "primaryColor": "#eef2f6", "primaryTextColor": "#1e293b", "primaryBorderColor": "#334155", "lineColor": "#b44527", "edgeLabelBackground": "#f9fafb", "tertiaryColor": "#ffffff"}, "flowchart": {"useMaxWidth": true, "htmlLabels": false, "nodeSpacing": 60, "rankSpacing": 100, "padding": 24, "curve": "basis"}} }%%
flowchart LR
  A["THE PARTY WINS"]
  B["IT CONTROLS THE JOBS<br/>AND THE CONTRACTS"]
  C["JOBS TO SUPPORTERS,<br/>CONTRACTS TO FRIENDS"]
  D["CAMPAIGN WORK, VOTES<br/>AND MONEY COME BACK"]
  A --> B
  B --> C
  C --> D
  D -- "so it wins again" --> A
  classDef step fill:#eef2f6,stroke:#334155,stroke-width:2px,color:#1e293b;
  class A,B,C,D step;

Federal civilian employees: 53,000 in 1871 → 256,000 in 1901.

The machine’s currency was the public job — and the job was public money.

Skowronek (1982, ch. 3). Officeholders were also “assessed” — taxed on their salaries to fund the campaigns that kept the party in office.

How would you break it?

Take the jobs out of the politicians’ hands.

In the United States that was civil service reform — the Pendleton Act of

Hiring by open competitive examination instead of by party.

It covered 11% of federal positions.
Enforcement, and any widening of that coverage, was left to the President.

Diego, twenty years later

Some jobs now need an exam. Diego could not have got one of those through the party. But the machine still fills the rest.

The exam did not take jobs off the machine. Total posts grew from 131,208 to 208,000 — and merit hiring staffed the new ones.

Reform grew alongside the machine instead of replacing it. Nineteen out of twenty new merit posts came from growth, not from a job the party gave up.

So what?

Civil service reform attacked the same mechanism as general incorporation: take a valuable favour out of politicians’ hands
But the Pendleton Act did it only partially — most federal jobs remained available for patronage
Over the next decades, merit rules spread unevenly, especially through city governments

That gives us something we can test. If public jobs stop depending on the party, should government officials behave differently?

Part IV — Did Reform Pay? Cities and Parallels

The test: patronage → merit rules → did government behaviour change?

Does it matter who does the hiring?

You run the city’s public works department. There is money for one project.

A park. Open in a year. Everybody can see it.
A sewer. Years of digging. Nobody notices until it works.

If you owe your job to the party, you need the party to win the next election. Build the park.

If you got your job by examination and can only be dismissed for cause, you will still be here when the sewer is finished. Now it is worth building.

So the prediction is: cities that hire by exam should spend a bigger share of the budget on the slow things — roads, sewers, water.

Rauch (1995) goes looking for it.

Where the evidence comes from

Rauch (1995) follows 144 cities, every year from 1902 to 1931 — every U.S. city above 30,000 people whose budget records were complete.

The map shows the 100 largest, the list the Census publishes by name. Rauch never prints his 144, so read this as the size class he drew from, not his exact sample.

Which reform actually protects the clerk?

Back to your public-works job. Three reforms spread through American cities. Ask one question of each:

after this reform, can you still be fired when the party loses?

Reform	Can you be fired?	So you build…
Civil service — hired by exam, dismissed only for cause	No	the sewer
City manager — one appointed boss at the top (Staunton, 1908)	The council can still dismiss him	unclear
Commission — elected commissioners run the departments (Galveston, 1901)	They face the voters themselves	the park

So only civil service should push spending toward the slow things.

And that is what makes it a test. If all three reforms raised infrastructure spending, the story would be wrong — it would just be “reform” doing something.

How the comparison works

Each city is compared with itself, before and after. Year effects strip out whatever was happening across the whole country that year.

But cities chose their own reforms. Rauch calls it a natural experiment; read it as comparison with controls, not randomisation.

Intuition: what we are hoping to see

Flat before. If the line were already climbing before adoption, reforming cities were on their way up anyway and the reform explains nothing.

Slow after. Sewers take years. An effect that appeared overnight would be suspicious, not reassuring.

Rauch does not publish this picture — he reports a single average effect. Hold this shape in mind and see whether his number is consistent with it.

And that is what he finds

Exactly the pattern the theory predicted — and the two comparison reforms rule out “any reform would have done it.”

Spending is sticky, so the long-run effect is about +2.2 points on a mean roads-and-sewers share of 13.8% — roughly a sixth more infrastructure spending.

The conclusion: protecting the official from the election cycle lengthened how long the city was willing to wait for a payoff. It was the content of the reform, not reform as such.

Reading the U.S. case onto today

What travels from the U.S. case?

U.S. case	The mechanism	Question to ask elsewhere
Charters reserved for a faction’s friends	Restricted entry creates a valuable political favour	Where does entering a market still depend on a political decision?
Public jobs rewarded party work	Patronage turns public employment into political currency	Who controls hiring and dismissal — and by what rules?
Merit rules covered only part of the state	Reform can leave important discretion untouched	Which positions remain politically controlled?
Reform took decades to spread	Rules survive when somebody has an interest in sustaining them	Who gains from reform — enough to defend it?

The point is not to look for another Gilded Age America.

It is to look for the same mechanisms in different institutions.

Exercise 2 · One public job

The city has one clerkship to fill.

All three applicants are qualified.

Applicant	Campaign workers they bring
Anna	2
Bruno	8
Carla	20

You are the party boss.

Who gets the job?

Of course you chose Carla

Why?

Not because she is the best clerk. Because the public job can buy you 20 campaign workers.

The job has become political currency.

Now change one rule

Same job. Same applicants. Same party boss.

But now applicants take an examination:

Applicant	Campaign workers	Exam score
Anna	2	92
Bruno	8	81
Carla	20	74

Who gets the job now?

What disappeared?

The job did not disappear.

The politician did not become more honest.

The party still wants to win.

But the boss can no longer say to Carla: “Support me and I will give you the job.”

The reform removed the politician’s discretion.

The ladder — ask, do not tell. Four questions, in this order.

“You are the party boss. Who gets the job?” — wait them out. Somebody says Carla. Do not move on yet.
“Why Carla?”

not because she is the best clerk.
The job is being paid for in campaign work.

“Now the exam scores. Who gets it?” — Anna, 92. Let the reversal land without commentary.
“So what disappeared?” — take wrong answers first, they are more useful:
- “The corruption disappeared” → no
- “Now it is fair” → careful. Anna won this post. The exam covers this desk and nothing else.
- “The job disappeared” → no, the city still needs a clerk.
Land it: the boss can no longer say “support me and I will give you the job.”

Then the callback, in one line: that is Part II’s move — remove the bargaining chip — applied to a single hire. And it is Pendleton at 11%: real where it reaches, and it did not reach far.

So what?

Corruption can be self-reinforcing, not just misconduct by powerful individuals
Give ordinary people control over a scarce public benefit and they work out how to use it politically — as you just did
One exit: remove the favour, so the bargain is worth nothing
The other: build a coalition that gains from new rules — slower, and partial
City evidence fits insulated bureaucracies investing longer

Conclusion

Two kinds. One takes money out of the state. The other uses the state’s favours to hold on to power
Nobody had to be greedy. Diego knocked on doors; the firm sent a donation
Punishing people did not work. Replace them and the arrangement carries on
Removing the favour did. Meet the published rules and nobody has to say yes
Where you cannot remove it, hire by exam — but that only reaches as far as it is applied
And it paid. Cities that hired by exam spent more on roads and sewers

The United States did not get cleaner because better people arrived. It got cleaner because there was less to sell.

It took fifty years, and it was never complete.

Next session: the midterm review — the Sessions 1–6 toolkit, applied to a case you have not seen.

References

Gibson, C. (1998). Population of the 100 largest cities and other urban places in the United States: 1790 to 1990 (Population Division Working Paper No. 27). U.S. Census Bureau. https://www.census.gov/library/working-papers/1998/demo/POP-twps0027.html

Glaeser, E. L., & Goldin, C. (2006). Corruption and reform: An introduction. In E. L. Glaeser & C. Goldin (Eds.), Corruption and reform: Lessons from America’s economic history (pp. 3–22). University of Chicago Press.

Rauch, J. E. (1995). Bureaucracy, infrastructure, and economic growth: Evidence from U.S. cities during the Progressive Era. American Economic Review, 85(4), 968–979.

Skowronek, S. (1982). Building a new American state: The expansion of national administrative capacities, 1877–1920. Cambridge University Press.

Underhill, I. (1914). Tammany Hall & 14th St. West, New York City [Photograph]. Library of Congress, Prints & Photographs Division (LC-USZ62-101734).

Wallis, J. J. (2006). The concept of systematic corruption in American history. In E. L. Glaeser & C. Goldin (Eds.), Corruption and reform: Lessons from America’s economic history (pp. 23–62). University of Chicago Press.