This article was originally published in The Wall Street Journal 

Exaggerated claims about political corruption from self-styled good-government groups are nothing new. Often they're made with no data to back them up. Perhaps more dangerous is when the claims are supported with bad data, yet reported faithfully by the media with little scrutiny.

So it is with the study released last month by a consortium led by the Center for Public Integrity, which concludes that of all 50 states, the one with the lowest risk of political corruption is . . . New Jersey.

Because other groups such as Common Cause and newspaper editorial boards including the New York Times and the Washington Post are already pointing to the "State Integrity Investigation" as evidence that more regulation of lobbying and campaign finance is needed, it's worth taking a critical look. The Garden State has many virtues, but a reputation for political integrity is not among them.

New Jersey's place at the top of the heap isn't the only curious conclusion reached in the study. Virginia—which in 2008 was deemed one of the best-governed states in the nation by the Pew Center on the States—earned an F grade, placing it with seven other states, including North and South Dakota, that allegedly have the greatest risk of political corruption.

These sorts of findings admit of only two explanations: Either New Jersey has gotten a bum rap in the past or something is very wrong with the State Integrity Investigation.

For starters, the study never actually defines what it means by corruption. Instead, the risk of corruption is defined by the presence or absence of certain laws—such as strict campaign-finance limits and lobbying disclosure—that good-government groups promote. But without a working definition of corruption, it is impossible to determine whether these sorts of reforms are the appropriate remedy.

Is regulation of state insurance commissions, for example, as important as lobbying disclosure as a means to combat corruption? Who knows? The study gives equal weight to both. Yet that's like assuming aspirin is as good as a herbal supplement because some people think both can cure headaches.

All of which leads to the biggest problem with the State Integrity Investigation—the dearth of evidence demonstrating that many of the promoted reforms, such as public input into legislative redistricting and registration of lobbyists, actually prevent corruption.

Despite years of effort by proponents of strict campaign-finance laws, there's no strong evidence that such laws affect either actual corruption or the public perception of corruption. Despite this absence of evidence, Virginia is penalized in the State Integrity Investigation because it has no campaign-finance limits (nor, it should be noted, any meaningful history of corruption).

The conclusions of the State Integrity Investigation are in conflict with other studies that have attempted to measure how well the states are governed. For example, the Pew study mentioned earlier, "Grading the States 2008," concluded that Utah, Virginia and Washington were the best-governed states in the country. The State Integrity Investigation gives them grades of D, F and B- respectively.

The Pew study may have methodological issues of its own, but if we are given two grades for a state—one that emphasizes actual governance (such as whether states use cost-benefit analysis for regulations, or engage in responsible spending practices) and another that measures whether states have laws that somebody thinks might do something about corruption—we'll go with the governance measure.

Fighting political corruption is an important goal, and many sensible policies, like the stricter enforcement of existing antibribery laws, can help achieve that goal. But others, like campaign-finance restrictions, impose heavy burdens on free speech. It isn't asking too much of good-government groups to come forward with evidence that their preferred policies actually reduce corruption. On that score, the State Integrity Investigation falls far short of the mark.