The problem of measuring corruption.
A big problem in the fight with corruption is that corruption is extremely hard to measure. Most organizations that study corruption in the developing countries use surveys and anecdotal evidence in their estimates of the issue.
While most experts in the world agree that corruption is a huge problem in the developing world and that it is a bigger problem in developing countries than it is in the developed countries, even organizations such as Transparency International and World Bank mostly use assumptions and anecdotal metrics in the creation of their indexes and reports.
Being able to collect personal evidence about corruption and assess its magnitude is not the same as being able to fully measure the extent of the problem. For instance, theoretically speaking, it is possible to make a claim that corruption is an effective way to incentivize a party directly and to pass funds directly in a peer-to-peer way in an economic system where the government and other mechanisms in the society are not functioning the way they are supposed to. Another way to theoretically look at corruption is to make an assumption that corruption occurs in the developing countries more than in the developed ones because there is a gap between the actual salaries of government officials in the developing countries and the market value of these salaries. In the developed countries, a government official can stay at a job for several years, and then, if his or her career is successful, quit working for the government and make a lot of money by writing books, being an expert on TV, giving interviews, getting a TV show, or becoming a lobbyist. This is what a lot of politicians have been doing in the developed countries. Some of the recent examples: even though Hillary Clinton lost an election, she then made a lucrative book deal and wrote a book. James Comey also wrote a book after Donald Trump fired him. In a way, working for a government in the developed countries provides people with opportunities even after they resign and these opportunities do not exist in the developing countries.
Typically, there are two issues with perceptions of corruption by residents of the developing countries. The first one is that people’s perceptions work fine when it comes to the inflation of prices, for example, the government saying that it is buying concrete for 1.5x, but they do not work fine when it comes to volume and amounts. Corrupt government officials know this well, which is why they find ways to steal money and inflate budgets so that people do not realize that it is happening. For example, in a contract that will have a price that is market-value and reasonable, but the volume would be inflated such as a road needing 15,000 cubic feet of concrete in reality and the contract being for 20,000 cubic feet of concrete. These are the details that a person with insufficient knowledge of the construction industry will not notice.
Because of this, actual corruption in a country and levels of corruption reported by surveys may be very different. Corruption has a lot of forms and many of these forms are very hard to notice and perceive.
Another issue with measuring corruption is that differences in the backgrounds of people, especially education, have a much higher importance in what people say about corruption that the actual levels of corruption itself. For example, Banerjee and Pande in 2009 have surveyed journalists, who often have a higher level of education than other segments of the population, about the levels of corruption and then compared the data from the surveys to the actual data on corruption. What they have found was that education and backgrounds played a critical role in what people are saying and even though the experiences of people with their government may be very similar, what people say in the surveys and how they perceive the reality would be very different.
While both of these issues are the micro-level issues, they do create problems for assessing macro-trends because macro-trend calculations consists of the micro-level surveys and when a significant percentage of those are wrong, then the macro numbers are also going to have discrepancies with reality.
Indonesia is a classic example of how this can happen. Muhammad Suharto was ruling the country for thirty-one years, from 1967 to 1998. When he resigned, many reports have claimed that the corruption in the country became worse, but this is not necessarily true. After the resignation of Suharto, the press in the country became much freer and journalists started writing more about the real issues in the country and they have also received access to more data than before. For this reasons, while the reports and press were claiming that the levels of corruption were going up, it is very possible that the levels have actually stayed the same or maybe even decreases, it is just that the society became more open compared to the past.