Two ways to compare economies

To compare two countries’ GDP you have to put them in the same currency — and there are two ways.

• Market exchange rates. China’s GDP is converted into dollars at the day’s rate: the price at which currencies are actually bought and sold.
• Purchasing power parity. Adjusts for what things cost in each country. A meal or a haircut costs less in China, so the same money buys more — and its economy looks bigger.

Catching up, depending on how you measure

What corruption is supposed to do to growth

China did all of that during decades when corruption was widespread. On the standard account, it should not have been able to.

%%{init: {"fontSize": 22, "htmlLabels": false, "flowchart": {"useMaxWidth": true, "nodeSpacing": 45, "rankSpacing": 55, "htmlLabels": false, "padding": 14, "curve": "basis"}, "themeVariables": {"fontSize": "22px", "fontFamily": "DM Sans, Segoe UI, sans-serif", "primaryColor": "#eef2f5", "primaryTextColor": "#1e293b", "primaryBorderColor": "#4a7c6f", "lineColor": "#334155", "secondaryColor": "#f6e7df", "tertiaryColor": "#f9fafb"}, "width": 1150, "height": 620}}%%
flowchart LR
  A["Officials can<br/>demand bribes"] --> B["Investing costs<br/>more and is<br/>less certain"]
  A --> D["Public money<br/>is diverted"]
  B --> C["Firms invest less<br/>and stay small"]
  D --> E["Worse roads,<br/>schools<br/>and courts"]
  C --> F["Slower<br/>growth"]
  E --> F
  classDef bad fill:#f6e7df,stroke:#b44527,stroke-width:2px,color:#1e293b;
  class F bad;

Author’s illustration of the standard argument. No units; conceptual only.

China did the opposite

• From 1978 to 2010 the economy grew roughly 9% a year
• Corruption did not fall as markets deepened — it changed shape
• Wedeman calls this China’s double paradox: growth and graft rose together

The simple cross-country relationship is weak

And China is not a lone exception.

Descriptive only. No controls for starting income, region or measurement error, and the base year is arbitrary. This motivates the puzzle; it does not test the corruption–growth hypothesis.

So what?

Counting corruption tells us how much. It does not tell us what kind.

• The standard argument predicts slower growth; China grew fastest while graft rose
• Across countries, the simple relationship is too weak to guide us either way
• Both point the same way: how much corruption is the wrong question
• Hong Kong ran the opposite experiment in the same region

Ang’s four types of corruption

Question 1. Who is involved — a clerk, or a powerful politician?

Question 2. What is the act — taking (one side loses), or trading (both expect to gain)?

Session 6: which way does the arrow point?

Wallis asked one question about any corrupt exchange.

%%{init: {"fontSize": 22, "htmlLabels": false, "flowchart": {"useMaxWidth": true, "nodeSpacing": 90, "rankSpacing": 240, "htmlLabels": false, "padding": 18, "curve": "basis"}, "themeVariables": {"fontSize": "22px", "fontFamily": "DM Sans, Segoe UI, sans-serif", "primaryColor": "#eef2f5", "primaryTextColor": "#1e293b", "primaryBorderColor": "#334155", "lineColor": "#b44527", "secondaryColor": "#f6e7df", "tertiaryColor": "#f9fafb"}, "width": 1150, "height": 520}}%%
flowchart LR
  A["PUBLIC OFFICE<br/>the power to decide"]
  B["ECONOMIC BENEFITS<br/>money, jobs, contracts"]
  A -- "VENAL<br/>office used for private gain" --> B
  B -- "SYSTEMATIC<br/>privileges used to sustain power" --> A
  classDef box fill:#eef2f5,stroke:#334155,stroke-width:2px,color:#1e293b;
  class A,B box;

Source: Wallis (2006), p. 25, in Glaeser & Goldin, eds.

Ang and Wallis on the same desk

The two frameworks ask different questions; the table below is only a rough alignment.

Access money always pays the official. It is systematic only where the privileges are handed out to keep a political coalition in power.

Ang (2020), ch. 1; Wallis (2006), p. 25, in Glaeser & Goldin, eds. The mapping is the author’s.

Why access money can ride alongside growth

The same four scenes, as a diagram. Where exactly was the corruption? Not in the building going up — in who was allowed to build, and what the official got for allowing it.

%%{init: {"fontSize": 22, "htmlLabels": false, "flowchart": {"useMaxWidth": true, "nodeSpacing": 45, "rankSpacing": 50, "htmlLabels": false, "padding": 14, "curve": "basis"}, "themeVariables": {"fontSize": "22px", "fontFamily": "DM Sans, Segoe UI, sans-serif", "primaryColor": "#e7efec", "primaryTextColor": "#1e293b", "primaryBorderColor": "#4a7c6f", "lineColor": "#334155", "secondaryColor": "#f6e7df", "tertiaryColor": "#f9fafb"}, "width": 1150, "height": 620}}%%
flowchart LR
  A["Scene 1<br/>only the state<br/>sells the land"] --> B["Scene 2<br/>cheap land for<br/>a share of the profits"]
  B --> C["Scene 3<br/>the towers go up<br/>the jobs are real"]
  C --> D["Scene 4<br/>no rivals bid<br/>the city is paid less"]
  classDef deal fill:#f6e7df,stroke:#b44527,stroke-width:3px,color:#1e293b;
  classDef cost fill:#f6e7df,stroke:#b44527,stroke-width:2px,color:#1e293b;
  class B deal;
  class D cost;

• Scene 2 promised him a share of the profits, not an envelope
• No towers, no profits, no share
• His career pointed the same way: officials were judged on growth

• Plain theft: the money leaves, nothing is built
• Access money: he is paid only if Scene 3 happens
• So the bribe and the building rise together

Author’s illustration of a mechanism Ang (2020) proposes, not an estimated causal effect.

What access money costs

Scene 3, ten years on. The towers are still standing. So is the debt that built them.

Four costs follow from the deal itself:

• Diego got the land through a connection, not because he builds well
• Across an economy, that means a lot of second-best firms doing the work
• The city was paid less than the land was worth, and that is public money
• What gets built depends on who knows whom rather than on what is needed

If it is legal, is it corruption?

Joseph Keppler, The Bosses of the Senate, Puck, 23 January 1889. U.S. Senate Collection; public domain.

Exercise 1 — Sort the bribe

Ang’s boxes

Source: Ang (2020), ch. 1.

Exercise 1 — where the cases go

1. Customs officer keeps part of the duties she collects — petty theft
2. Restaurant pays a clerk so the licence arrives this week — speed money
3. Minister moves pension money into a private account — grand theft
4. Developer funds the mayor’s project; the city then rezones its land — access money

Source: Ang (2020), ch. 1.

Hong Kong is not mainland China

Two histories, one coastline

Founding date: ICAC, History of the ICAC.

Hong Kong before the ICAC

• Ambulance crews asked for “tea money” before picking up a sick person
• In hospital, a tip bought you a bedpan or a glass of water
• Firemen were paid twice: once to turn the hose on, once to turn it off
• None of this was freelance. The police ran it as a business — syndicated corruption: set prices, collected on a schedule, shared upwards through the ranks

An agency placed outside the police

ICAC headquarters, North Point, Hong Kong. Photo: Citobun, CC BY-SA 4.0, via Wikimedia Commons. The Commissioner reports to the Chief Executive alone, and to the Governor before 1997 — not to the police, not to a ministry.

The ICAC attacks the loop at three points

%%{init: {"fontSize": 22, "htmlLabels": false, "flowchart": {"useMaxWidth": true, "nodeSpacing": 40, "rankSpacing": 55, "htmlLabels": false, "padding": 14, "curve": "basis"}, "themeVariables": {"fontSize": "22px", "fontFamily": "DM Sans, Segoe UI, sans-serif", "primaryColor": "#e7efec", "primaryTextColor": "#1e293b", "primaryBorderColor": "#4a7c6f", "lineColor": "#334155", "secondaryColor": "#f6e7df", "tertiaryColor": "#f9fafb"}, "width": 1180, "height": 640}}%%
flowchart LR
  I["ICAC<br/>1974"] --> O["Deterrence:<br/>investigate<br/>and prosecute"]
  I --> P["Prevention:<br/>redesign<br/>procedures"]
  I --> C["Education:<br/>persuade and<br/>take reports"]
  O --> R1["Raises the risk<br/>of taking"]
  P --> R2["Removes the<br/>opportunity"]
  C --> R3["Changes what<br/>people expect"]
  R1 --> N["Clean becomes<br/>the new normal"]
  R2 --> N
  R3 --> N
  classDef goal fill:#e7efec,stroke:#4a7c6f,stroke-width:3px,color:#1e293b;
  class N goal;

Author’s illustration. Manion (2004), ch. 2: reform needs all three prongs together.

▶ Hong Kong — worked

1. Why education?
   • The loop holds because everyone expects everyone else to pay
   • Arrests shift that expectation only if people hear about them
   • Education is how a prosecution becomes an expectation
   • The three prongs work as one
2. Why the joke can be sold now
   • Nobody doubts the threat behind it any more
   • Being summoned has to be real and routine before it can be funny
   • The open day is not time off from the work — it is the education prong
3. Reading the source
   • The ranking is a perceptions index, with all of Part I’s problems
   • No counterfactual: what else changed in Hong Kong after 1974?
   • A mainland broadcaster in 2014 had its own reason to tell this story — hold that against Part IV

The history segment of this report (2:52-4:36) covers Godber and the 1960s salary problem.

Making the new expectation credible

• Independence. The Commissioner answered to the Governor directly, not to the police
• Permanence. A standing agency with its own staff and budget
• Visibility. Senior officers prosecuted, publicly and repeatedly
• A line under the past. Police protest in October 1977
• Partial amnesty: offences committed before 1 January 1977 not pursued
• Those already under investigation, wanted, or abroad were excluded

Sources: Manion (2004), ch. 2; ICAC official history. The amnesty closed the past once — not repeatedly.

Two routes to clean government

Session 6 ended on a sentence worth repeating: the United States did not get cleaner because better people arrived — it got cleaner because there was less to sell.

Sources: Manion (2004), ch. 6; Wallis (2006), p. 50, in Glaeser & Goldin, eds., Corruption and Reform.

But was it really the ICAC?

• The ICAC arrived in the middle of a much wider reform drive
• Hong Kong was small, rich and urban, not a continent
• Rising incomes made a civil-service job more costly to lose
• Courts and newspapers operated with real independence
• Manion shows the design was coherent; she cannot randomise it
• Read Hong Kong as a strong case, not a controlled experiment

Sources: Manion (2004), ch. 2; ICAC official history. These rival explanations are not tested here — that is the point.

Exercise 2 — Would an ICAC work here?

So what?

Manion’s answer for Hong Kong: it did not punish harder. It made punishment credible, and designed around it.

Two designs, side by side

Sources: Manion (2004), chs. 5–6; CCDI. Since 2018 a National Supervisory Commission extends the mainland’s reach to all public employees. Caution: these bodies sit inside very different political systems. This is a structured comparison, not a controlled experiment.

Xi’s campaign: real, and very large

Sources: CCDI press conference, June 2022 (as reported); court judgment in the Zhou Yongkang case, 11 June 2015.

What a political connection was worth

Diego’s deal, many times over. Nobody can observe a secret profit share. But the state records every land sale, and Chen and Kung can tell which buyers were linked to the families of Politburo members — “princelings”. The favouritism turns up in the price, with nobody having to confess.

Source: Chen and Kung (2019), Quarterly Journal of Economics; over one million transactions, 2004–2016. Prices are recorded by the state, not self-reported in a survey.

Why would an official give that discount?

The buyer is not a stranger. A “princeling” firm is owned by the family of a Politburo member — a son, a son-in-law.

• Nobody owns urban land in China — the state does, and has since 1982
• What is sold is a use right for 40 to 70 years, granted by the local government
• Say a parcel is worth $100 million; the princeling firm pays about $42 million
• Its owners are instantly $58 million richer
• Those owners are a national leader’s children
• That leader sits where promotions to national office are decided

Illustrative figures, using the midpoint of the discount estimated by Chen and Kung (2019). Urban land is state-owned under the 1982 Constitution; use-right terms come from the 1990 Interim Regulations.

Enforcement that is real, and also selective

Did the campaign change anything? Chen and Kung watch the same land prices before and after 2012.

• A province was “reached” when a central inspection team arrived, or when Xi appointed its new party secretary
• In those provinces the princeling discount shrank by about a third to two-fifths
• Princelings still paid less than everyone else, but much closer to the ordinary price — so the enforcement was real

So what? Five things to take away

• Ask what kind of corruption a country has, not only how much
• Access money can live alongside growth; theft and extortion cannot
• The bill still arrives: the best builder loses, and the city is underpaid
• Corruption lasts because people expect it, and in 1973 nobody could refuse alone
• Clean government came from changing structure, not from trying harder

References

Ang, Y. Y. (2020). China’s gilded age: The paradox of economic boom and vast corruption. Cambridge University Press.

Chen, T., & Kung, J. K.-S. (2019). Busting the “princelings”: The campaign against corruption in China’s primary land market. The Quarterly Journal of Economics, 134(1), 185–226. https://doi.org/10.1093/qje/qjy027

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.

Global Times. (2022, June 30). China punishes 4.7 million people in decade-long anti-graft campaign. https://www.globaltimes.cn/page/202206/1269498.shtml

References (continued)

Independent Commission Against Corruption. (n.d.). History of the ICAC. Government of the Hong Kong Special Administrative Region. https://www.icac.org.hk/en/about/history/index.html

Manion, M. (2004). Corruption by design: Building clean government in mainland China and Hong Kong. Harvard University Press.

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.

Wedeman, A. (2012). Double paradox: Rapid growth and rising corruption in China. Cornell University Press.