New Zealand had the top scores for anticorruption in this year’s Corruption Perceptions Index. Pictured above is Prime Minister Jacinda Ardern. (SeongJoon Cho/Bloomberg News)

Curious about the world’s most — and least — corrupt countries? On Wednesday Transparency International (TI), a well-known NGO devoted to global anti-corruption, released its annual scorecard.

TI’s Corruption Perceptions Index (CPI) shows the United States now sits in 16th place (of 180 territories), with a score of 75 out of 100. That’s a small improvement on the 2016 index in which the United States was 18th with a score of 74.

Yet barely a day seems to go by without new corruption allegations involving the Trump administration — including conflicts of interest among top officials and obstruction of justice claims against the president himself. How does that square with this latest CPI report?

How the CPI measures corruption

The CPI is probably the best-known measure of global corruption. The 2017 CPI takes into account 10 surveys and assessments from a range of institutions. The data come predominantly from surveys of business people around the globe.

TI amalgamates that data with assessments by a range of independent analysts to generate each country’s score. A score closer to 100 indicates the country is doing a decent job at preventing public sector corruption. A score below 50 suggests the country has a serious problem, while a score that falls below 30 is an indication that corruption is systematic and systemic within that country. Little more than 30 percent of countries score above 50, while around 3 in 10 score below 30. A worrying picture.

What should we read into the apparent improvement in the U.S. ranking? In truth, not much. Trump’s supporters may well use this as evidence that all the apparent “fake news” and scaremongering is covering up a more positive story.

In the real world of global politics, these sorts of indexes should be used with caution. As I illustrate in a new book, aggregate indicators of corruption are, at best, a very blunt tool for assessing what is going on in the world and, at worst, nothing more than out-and-out guesswork. I say that for four reasons:

1) Indexes generally measure perceptions of corruption. Even TI urges caution in reading too much into its findings. The CPI authors are quick to acknowledge that this table doesn’t look at actual levels of corruption but, rather, perceived levels of corruption. There can be a significant difference between those two things. One person’s bribe often can be another’s deal; one person’s perception of a swamp that needs draining might be another’s messy way of getting politics done.

2) The ranking is just a number. How can we boil the complex and largely clandestine process of corruption down to one number? China and Trinidad and Tobago, for example, each scored 41 (and are subsequently tied in 77th place), but does it make sense to argue that the complex corruption challenges of fast-developing China are on par with two islands in the Caribbean with a combined population of about 1.3 million?

3) The survey methodology is based on other surveys. The CPI scores rely on data from various surveys, which may have been undertaken for different reasons. TI has effectively developed a method to group the data in a poll of polls, but the number of data sources ranges from three (in the cases of places such as Dominica, Grenada and Barbados) to 10 (in the likes of Poland, South Korea and Estonia). Efforts to create a rigorous methodology notwithstanding, many scholars severely doubt that this is a nuanced enough database to create a meaningful table of corruption.

4) It takes time for real-world events to filter into the indexes. That’s often the case and in two distinct ways. In a practical sense, the data come from surveys that were conducted in the months or even years running up to the publication of that year’s CPI. The current table is therefore at best a snapshot of what was happening at a point in the past, not today.

In a psychological sense, it also takes time for corruption developments — both good and bad — to be processed by those responding to surveys like these. In the immediate aftermath of a very high-profile parliamentarians’ expenses scandal in Britain in the late 2000s, the United Kingdom’s scores in the CPI went down. Yet the British Parliament took quick action, and parliamentarians got the message to avoid risking any further misdemeanors. It took years for the U.K. scores to start improving again.

China also saw its score dip after the launch of Xi Jinping’s extensive anticorruption drive in 2012. Only now are we seeing China’s CPI score recover. This suggests that acknowledging and then fighting corruption can lead to worse scores rather than better ones, at least initially.

Who are the real anticorruption stars?

Although New Zealand came out on top in the 2017 index, it is the Nordic countries that continue year on year to be star performers. The reasons seem straightforward. The Nordic nations — Denmark, Finland, Iceland, Norway and Sweden — certainly have their scandals, but they remain open and transparent democracies with, importantly, a free press. Their quality of governance, no matter how you measure it, is excellent.

Furthermore, and perhaps counterintuitively, these governments spend lots of taxpayers’ money; in 2016 the government expenditure to GDP ratio in Denmark was 57.1 percent. In other words, the Danish state spends a lot of money on public services, and yet that doesn’t lead to more corruption (as many argue it could).

Perhaps there’s a broader takeaway here for other countries, including the United States. Those who focus on cutting the role of the state might want to think beyond the budget numbers and consider what the state does — and make sure it does its job well.

Trump watchers curious about the effects of a Trump presidency on perceptions of corruption in the United States might not find many insights here, but stay tuned for future CPI reports in the years ahead. That’s where any fallout from the Trump presidency is likely to be revealed.

Dan Hough is a professor of politics at the University of Sussex and director of the Sussex Centre for the Study of Corruption.