THE latest Corruption Perception Index (CPI) shows Pakistan has improved its rank compared to last yearís (although its overall rank is still 116 out of 176). Politicians in power have been claiming the credit. Some provincial governments even spent money on advertisements highlighting the improvement. Itís certainly heartening to see imp*roved rankings of our country on any index. There are, however, a few qualifying issues related to the CPI that we should consider.

First, as the name implies, the index is based on perceptions. Good decisions on sensitive issues cannot be made on perceptions alone. Imagine yourself selecting a heart surgeon based solely on othersí perceptions. It may require more due diligence. Are the people forming perceptions biased? Are they experts? Were they ever his patients? Do they work with him or for him? Do any of them have incentives to form those opinions?

Transparency International aggregates data for the CPI from sources like World Bank, World Economic Forum, etc. Our government is a client of some of these organisations. Therefore, there certainly is a conflict of interest. This may cloud the judgement of some of the people involved.

The socioeconomic background of people offering their perceptions can also affect their judgements. Itís hard to see how an affluent, educated citizen who rarely interacts with less privileged citizens or the public institutions of the country heís measuring, can correctly assess the countryís corruption and sum it up in a number. People perceiving corruption in a certain country can also be subject to cognitive biases ó eg the bandwagon effect (if everyone else is saying itís not a corrupt country, then I will say that also) ó or form their perceptions based on certain stereotypes or isolated incidents.

How accurate is the CPI in measuring corruption?

Further, itís hard to imagine that a complex social phenomenon such as corruption can be reduced to a single number. The beauty is in the details and the CPI score doesnít give us much to ponder. Does it represent the frequency of corrupt activities or the magnitude of corruption? Does it tell us more about petty corruption or about mega corruption? Does it tell us anything about corruption in the private sector?

It definitely doesnít tell us which provinces or cities are more corrupt and which less so. Has staff truancy actually decreased in certain departments? Has it become easier for people to receive quality healthcare and impartial justice?

Additionally, corruption is clandestine in nature. Can a certain number indicate the true scale of corruption when it doesnít capture transactions that happen under the table? Our huge informal economy also makes it hard to capture the actual extent of corruption.

Sometimes, corruption today is discovered only years later; this phenomenon cannot practically be captured by indexes like CPI. This happened when Chile was ruled by Pinochet and the country got decent CPI rankings. Many corruption scandals erupted later on, when it became clear that Pinochet had garnered much wealth through corrupt activities during the time Chile was given good CPI rankings.

Corruption is usually defined as the use of public office for private gain. Less use of public office for private gain should result in improved quality of government. So if we are less corrupt now, has the quality of government actually improved? A good measure of a high-performing government and bureaucracy is how much tax revenue is being collected. Looking at the Tax-to-GDP ratios of different countries, itís surprising that some countries that are considered more corrupt (as per CPI) than us have better tax-to-GDP ratios. This makes one wonder how CPI takes into account the quality of a government.

Although the media may have its own biases and, in this competitive world, is caught up in a race for ratings, a free media does play an important role in forming perceptions. In the absence of free media, all you can rely on is gossip, anecdotes and your interaction with select individuals. This cannot lead to perceptions representative of the actual scale of corruption. Therefore, a free media is important for forming perceptions closer to reality.

However, we see countries like the UAE, Saudi Arabia, Tunisia and Cuba ó where itís known that dissent and local media are suppressed ó performing well on the CPI. It makes one wonder how societies that are known to curb dissent and media freedom rank high on an index based totally on perception.

All in all, the CPI is a corruption index based on perceptions of people from various organisations and social strata. I think we should celebrate that Pakistan has performed marginally well this year. However, we should take the rankings with a grain of salt rather than use them as a PR tool to claim political victories.