My AFR oped today is on the economics of corruption. Full text over the fold.

Cultures of Corruption, Australian Financial Review, 11 August 2009

Leisha Harvey was jailed for a year for using her official credit card for personal expenses, including taking her husband to enjoy the Grand Prix. Brian Austin was imprisoned for 15 months for misappropriation. Merri Rose spent 18 months behind bars for attempting to blackmail the Premier. Gordon Nuttall was jailed for seven years for corruptly accepting $360,000 from two businessmen.

What do all these cases have in common? They all involve Queensland politicians. Indeed, of the 16 state politicians who have been convicted of crimes in the past 20 years, 8 were from Queensland. According to Tony Fitzgerald, whose 1987-89 inquiry was meant to have ended official corruption in Queensland: “Access can now be purchased, patronage is dispensed, mates and supporters are appointed and retired politicians exploit their political connections to obtain success fees for deals between business and government.”

Of course, other states have had political corruption scandals of their own. New South Wales corrective services minister Rex Jackson’s cash-for-early-release scheme was nothing if not audacious. And Western Australia deserves a special mention for having had a Liberal Party premier and a Labor Party premier jailed in the 1990s. But for convictions, Queensland tops them both.

True, just 8 convicted politicians is a crude basis on which to attach the tag ‘most corrupt state’. Another approach is search the national media, and estimate what share of the approximately 100,000 state political stories published in the last 14 years mention corruption. On this metric, Western Australia and Queensland are the most corrupt states (with 10% and 8% respectively of political news items mentioning corruption), while the ACT and NT rank as the least corrupt (with 6% of political news items mentioning corruption).

What explains why some states are more corrupt than others? Since Gary Becker’s seminal work in 1968, most economists have regarded crime as a function of the expected benefits and the expected costs. When the potential gains from corruption are high, we should expect to see more of it. Conversely, when the penalties are large, or detection is likely, we should expect to see less corruption.

This suggests that states with more dispersed populations might experience more corruption, since a more dispersed population means fewer people in the capital city to keep an eye on the politicians. As one of the least urban states in Australia, it may have been harder for Queenslanders to prevent their MPs from misbehaving. Another factor could be the level of development. In the United States, Mississippi and Louisiana are among the states with the highest proportion of public officials convicted of corruption, while New Hampshire and Oregon are among the least corrupt. The fact that Queensland incomes have historically been below the national average may be relevant here.

But economists are also beginning to recognise the importance of a ‘culture of corruption’. When everyone around is thumbing their nose at the law, an otherwise scrupulous politician may be more tempted to do likewise. Or as Governor Willie Stark says in All the Kings’ Men, “dirt’s a funny thing. Some of it rubs off on everybody.”

Intriguing evidence on cultures of corruption comes from a study by Ray Fisman (Columbia University) and Ted Miguel (University of California, Berkeley), who observe that diplomats from countries with high corruption ratings (such as Egypt and Chad) are more likely to exploit their diplomatic immunity to accumulate unpaid parking tickets in New York City. Holding constant a country’s level of development, more corruption back home means more unpaid parking tickets in Manhattan.

Partly because of these cultural norms, fighting corruption is no quick process. But the most promising solution is to raise the chance of detection. In the words of former US Supreme Court Justice Louis Brandeis: “sunshine is the best disinfectant”.

In Australia, independent corruption-monitoring bodies have doubtless helped. But perhaps just as important is the role of an active and independent media. Just as the Crédit Mobilier and Teapot Dome scandals were both brought to light by the US media, so too the Courier Mail and Four Corners helped expose Queensland police corruption in the 1980s.

In the Internet age, the media should also press for more data to be available online. Political donations could be posted quarterly rather than annually. Lobbyists ought to report on their fees (as occurs in the US), not merely their client list. And political parties’ ‘business liaison’ programs should be fully reported. If something dodgy is going on, posting the jigsaw online raises the odds of someone piecing it together. Perhaps the “Sunshine State” could lead the way?

Andrew Leigh is an economist in the Research School of Social Sciences at the Australian National University.

The data on politicians who have been convicted of crimes was based on a Wikipedia list of Australian politicians convicted of crimes. It was incomplete, so my research assistant, Susanne Schmidt, spent a day searching media databases and the web to update it. We think we’ve found all the convictions in the past 20 years, but naturally if you know of more, please update Wikipedia.

For anyone who’s interested, details of the citations exercise are below.

Results of citation analysis

Note: The citations analysis was based on a Factiva search of the Australian and the Australian Financial Review over the period 1996-2009 (the full availability in the database). In about 2004, the AFR’s Factiva coverage switched from full text to summaries.

All other facts and pieces of research are hyperlinked in the article.