Just how bad is corruption in China?

In a new paper, my colleague Carlos Ramirez has the scoop:

This paper compares corruption in China over the past 15 years with corruption in the U.S. between 1870 and 1930, periods that are roughly comparable in terms of real income per capita. Corruption indicators for both countries and both periods are constructed by tracking corruption news in prominent U.S. newspapers. Several robustness checks confirm the reliability of the constructed corruption indices for both countries. The comparison indicates that corruption in the U.S. in the early 1870s — when it’s real income per capita was about $2,800 (in 2005 dollars) — was 7 to 9 times higher than China’s corruption level in 1996, the corresponding year in terms of income per capita. By the time the U.S. reached $7,500 in 1928 — approximately equivalent to China’s real income per capita in 2009 — corruption was similar in both countries. The findings imply that, while corruption in China is an issue that merits attention, it is not at alarmingly high levels, compared to the U.S. historical experience. The paper further argues that the corruption and development experiences of both the U.S. and China appear to be consistent with the “life-cycle” theory of corruption — rising at the early stages of development, and declining after modernization has taken place. Hence, as China continues its development process, corruption will likely decline.


'Corruption indicators for both countries and both periods are constructed by tracking corruption news in prominent U.S. newspapers.'

And to think the term 'yellow journalism' stems from that earlier American time period too. ('The term originated during the American Gilded Age of the late nineteenth century with the circulation battles between Joseph Pulitzer's New York World and William Randolph Hearst's New York Journal. The battle peaked from 1895 to about 1898, and historical usage often refers specifically to this period. Both papers were accused by critics of sensationalizing the news in order to drive up circulation, although the newspapers did serious reporting as well.' - http://en.wikipedia.org/wiki/Yellow_journalism And do note Politzer's name there. when thinking about 'serious reporting in major American newspapers.' )

'The comparison indicates that corruption in the U.S. in the early 1870s — when it’s real income per capita was about $2,800 (in 2005 dollars) — was 7 to 9 times higher than China’s corruption level in 1996, the corresponding year in terms of income per capita.'

So, we are comparing a country that had just finished a civil war (which just happened to kill more of its citizens than any other war before or since in its history), having just released a significant fraction of human beings owned as chattel, with another country that was on the verge of reacquiring one of the more valuable pieces of well managed real estate in the world?

Even more interesting, considering just how much anti-corruption law is a modern invention (what food adulteration laws looked like in the U.S. in 1890 is not something people need concern themselves with - the morphine in your dropsy medicine was not regulated, after all).

'Hence, as China continues its development process, corruption will likely decline. '

Or more closely approximate the current American model, where massive fraud (documented fraud, it must be noted, with a paper trail) is something that no one bothers much about, while classifying those institutions which had engaged in such massive and ongoing fraud as too big too fail. Which isn't corruption, of course, it is just the American Way, apparently.

Did they account for the 1st Amendment of the U.S. Constitution, and lack of it in China?

Upon beginning to skim the article, yes, he does.

"In particular, following the Glaeser and Goldin (2004)(7) methodology, I first document the dynamics of corruption in the U.S. using newspaper evidence from 1870 to 1930. Next, I develop a similar time series evidence for China for the 1990 to 2011 period, using prominent U.S. newspapers as a source."

and he had the foresight to address these comments before we had to chance to write them (pg. 7-8).

And this methodology doesn't raise at least a warning flag? - 'using newspaper evidence from 1870 to 1930. Next, I develop a similar time series evidence for China for the 1990 to 2011 period, using prominent U.S. newspapers as a source'

Not simply because national reporting is influenced by national perspective (something Americans seem to believe only occurs when other nations report about the U.S., though never in reverse), but because China jails foreign journalists when they commit 'espionage,' such as reporting information the Chinese do not want published?

Then, for extra credit, ask yourself how many American journalists in China can actually speak and read Chinese, so as to be at least aware of what it is their 'hosts' are trying to keep them from becoming aware of. It is a sad joke that far too often, American foreign journalists have no knowledge of the actual language(s) of the country(ies) they are reporting from, making them even easier to manipulate. And this is a major reason why so much 'American' foreign news from the EU/eurozone is just warmed over British reporting, generally with all the UK bias apparent to anyone who bothers to actually read the British reporting first.

I don't want to sound like an American apologist, and the conclusion very well be correct, but the methodology certainly looks questionable.

As prior_approval suggests: isn't it a bit odd to assume equality in journalistic standards? I haven't read the paper yet, but the conclusion seems to be contingent on the assumption that reporting in modern China is comparable to reporting in the US. Given differences in 1) social norms 2) the way print media functions as a whole and 3) restrictions on free speech, how can this comparison really be considered valid?

comparable to reporting in the US circa 1900*

Now and then, the Chinese government hangs an official to encourage the others.

Actually they usually shoot them. Although they are moving to lethal injection.

If someone else has read the paper I don't really have time, but I would be interested to know how well he can really expect to account for:
1. The fact that corruption is probably easier to hide in an economy with a lot of state-owned enterprises and no freedom of speech (I know he accounts for the lack of freedom of speech but when combined with the prevalence of SOEs this might be a bigger problem then normal)
2. How large the impacts of these instances of corruption are
3. How many articles there are going to be on corruption in a US newspaper reporting on US news compared to foreign news (I would assume they compare stories of corruption to overall news stories about the area, but stories on corruption are going to have more draw if they are about local corruption which makes the local population feel scandalized)
4. how related instances of large-scale corruption are to corruption at a more micro-level (if every license requires a bribe, this won't make it into the news, though it might have a larger negative effect on growth)

When combined with the other reasons listed above, it just seems like there are so many reasons that the number would be higher for the US that it's hard for me to be surprised by the results.

methinks the size of the public sector in China today and the US in the 1920s aren't really comparable.

It's one thing to have a corrupt official making money on side deals, favorable decisions and good old bribery. It's another thing when the local government is essentially the local economy and decides all the major economic decisions. Corrupt U.S. politicians became wealthy, but they were still small fish compared to the titans of industry. In China, the corrupt politicians are the titans.

(Apologies for wall of text)

I think it's rather naive to compare corruption in the US and China by only looking at US newspapers, as his methodology suggests. Foreign journalists are bound to have less access than Chinese journalists. In some of the more intense cases, foreign journalists are flat out refused entry into certain parts of China. For example, it's currently a huge ordeal to get into Tibet, even for travel.

I assume the author omitted Chinese papers in his investigation due to state influence, but that sort of tells its own story on corruption anyway. You might argue that US commercial newspapers also had (and continue to have) strong political forces influencing their coverage, but they also weren't forbidden (at least to my knowledge) from reporting stories that take place outside their localities. Chinese papers ARE forbidden from doing investigative reports outside of their home location, and are only permitted to publish Xinhua wires on major issues happening outside the home province or city. At least in the 1890s, the New York Times could theoretically run a story on corruption in Philadelphia. Can't do that in China.

I think the Corruption Perceptions Index also merits a review in terms of credibility. The index measures perceptions, not actual corruption (which by its nature is almost impossible to measure). If you've never personally paid a bribe, you might think the country is relatively clean, regardless of what officials are doing with your tax money behind your back. Or based on cultural nuances, you might see nothing wrong with giving a teacher a large monetary gift to ensure your child gets into a preferred university. Obviously that would be a problem in the US, but it's not always considered outlandish in China. Officials drive cars and wear watches that far, far outweigh what someone with their official salary could possibly afford, and this is relatively well accepted. In some countries that have a high corruption ranking, an official driving a Porsche would be a huge red flag, because there is more civic activism. That doesn't mean the "corrupt" country is fundamentally more crooked than China, just that the public perception is that officials are genuinely more corrupt. I've seen many a Porsche Cayenne in China with military plates, but people don't seem to get worked up over where those soldiers are getting their income from.

I would also add that you have to question the wisdom of a Chinese farmer answering these sorts of sensitive questions to foreign agencies running corruption perception surveys. There is a chance, maybe only 1%, that their responses could get them into big trouble. In other countries, you can talk about corrupt officials all day, and your freedom isn't in danger. In China, giving a "bad" interview to a foreign reporter has landed more than a few people in jail. So what is their incentive to answer the questions accurately? I would guess that in the Philippines or Greece - both "more corrupt" than China - the threat of being thrown in jail over calling out a corrupt official is probably less than in China.

I honestly don't know which country is the most corrupt. I would say the US ranks relatively poorly in the CPI because most Americans inherently distrust their government (more so than other nations). For China specifically, I would also look at the Gini index published by their government yesterday: 0.61. This makes China one of the most unequal countries on earth. Corruption is likely part of that.

China grew as much in the last 15 years as the US did in 60? And those 60 years of growth are generally classed as exceptional weren't they? There is not great stagnation.

I believe Scott Sumner argues a variant of this.

I believe the "Great Stagnation" hypothesis is for fully developed economies and a developing economy rapidly catching up with the developed economies doesn't disprove the thesis.

highly sceptical of relevance of a study like this - way too many variables, the biggest of which being population size and demographics in general, divergent historical factors influencing the behaviors of that citizenry and of course a sharp differentiation viz political systems and rule of law - just way too many variables that can't be effectively accounted for - might as well compare corruption in China to that in the court of louis XIV - or post Gorbachev Russia - or hell how bout nazi germany, the last time an autocracy embraced 'capitalism'.

So over a time series chosen to match GDP per capita, corruption reports *multiplied 7-9 times faster* in China.
And this leads us to predict that corruption is about to decline in China.
Just like it already is declining in Russia, we must assume, since Russia is far ahead of China on the "life-cycle" curve.

Yeah but we got The Godfather and Gangs Of New York--am I really ever gonna wanna watch a movie about Bo Xilai? Please.

You already have. Its called The Godfather.

1) The 1870s was not as economically optimized a time. Our modern/high-tech/globalized world is much much more vulnerable to crashes and their effects.

2) They're using one data point to say "China's not in trouble." I suppose I shouldn't worry about playing in traffic because someone else I know does it 7-9 times as frequently as I do.

3) During the 1870s, there was a gigantic economic crash. Many argue that this one was considerably worse than the Great Depression.

4) Doesn't China have a repressed media? Wouldn't that have an effect on how much corruption is "reported"?

5) This metric is squeezing everything into one metric of "corruption"--what about qualitative differences in types of corruption?

6) I was going to ask nerdier statistical questions, but I'm really suspicious of any methodology that's relying on correlations between words in newspaper articles. Is there any evidence whatsoever that this can be used for prediction?

China has just about the right amount of corruption or somewhat less, as compared to the U.S.? I predict this paper will be very popular in China, probably get a mention soon on the front page of People's Daily.

Perhaps the authors could follow-up with a paper on the leading role of the Party in guiding China through this period of rapid development. Maybe they could make an analogy with the role of ARPA in developing computer technology and the Internet. They could become China's favorite Western economists.

Corruption is notoriously difficult to measure, and using reports from newspapers in one country of corruption in another, seems to me to be specious thesis methodology at best.

China has become more salient due to the increased trade the US has with it, which increases US newspaper interest in corruption and the US businesses involvement in it. But I don't think it is viable to use that as an indicator for the increases or decreases in corruption. During the cold war, there were probably more reports on Russian corruption, salience.

Corruption is not entirely political, but much of it is, so looking at (state-controlled) Chinese newspaper reports would be a better indicator of the willingness to acknowledge and combat corruption and improve the quality of institutions - Indicating long-term reduction in corruption.

I think the argument that corruption in china will see a downward trend in the future is correct to some degree at the moment, but given the one-party-system, it is a much different beast than corruption in 1870's america.

Yeah I was hoping for some really creative measure of corruption. What about like changes in wealth of cousins of government officials pre v. post appointment?

I would think the best measure of corruption would be something like:

Does wealth and economic power precede political clout and connections, or does political clout and connections precede wealth and economic power?

I don't have access to the full article, so hope that the 'historical context' mentioned in the abstract is expanded upon. The levels of corruption in the US during the latter half of Prohibition ('20 - '33) is not a 'normalized' time period for corruption in this country's history - it is a time of exceptionally high corruption.

And for that matter the period from 1865-1880 (post Civil War) was one of renowned corruption also.

Has the study's author paid any attention to the history of newspapers? 1870-1930 was probably the golden age of American journalism. Huge numbers of newspapers, high quality, constant circulation wars, the development of press syndicates. You had the coexistence of newspapers that served as party organs (Ever notice how many newspapers are titled something like "The Arkansas Democrat"? There's a reason) with sensationalistic tabloids, with muckraking exposes.

If you go to your city library, you'll probably find that it had four or five times as many newspapers in 1890 as today, with a much larger focus on local politics and scandals. It would be ludicrous to compare the corruption beat of 1890 to today even if it were the same newspaper covering the same city. Comparing it to their coverage of Chinese corruption is just insane.

The conclusion is wrong.
No official newspapers or websites reported corruption.
Corruption is clearly higher in China.

A lot of golf courses were built in the U.S. in the 1920s and in China in the 2000s. My impression is that acquiring land for golf courses was a lot less corrupt in America in the 1920s.

Many flaws in this paper's approach are obvious and are covered above, so let me see if I can muster a small defence for it. One problem I have in general with people who study economics (and other subjects) is they too often look to each inquiry as a potential proof. But sometimes all an inquiry seeks is to shine some light on a previously unseen portion of a subject. I have not read this article but I assume that is what the author is attempting.

We all believe we know that corruption in China is epidemic. But that tells us very little. This flawed comparative approach at least partly addresses the issues of how relatively corrupt China is, even if it is across centuries, languages, cultures, legal systems etc.. There is little danger in inquiring whether the current level of corruption (as reported in American papers) is unique to China or perhaps just a stage of development. As long as one does not leap to any conclusions.

Comments for this post are closed