THE BLOG

From GDP To Crime, Statistics Are Unreliable Storytellers

06/07/2016 7:47 AM IST | Updated 15/07/2016 8:27 AM IST
NEW! HIGHLIGHT AND SHARE
Highlight text to share via Facebook and Twitter
John Lamb via Getty Images
Tablet computer with financial data.

Denmark, the happiest country in the world also happens to be the second-largest consumer of anti-depressants.

Statistics can be extremely deceptive to the unaware. Depending on the statistic one chooses to use or ignore, contrasting images can be drawn before the public.

What can you infer from crime rate statistics?

Kerala, the southern state of India, offers several interesting insights. Although known for its excellent social indicators and governance, the state has the highest crime rate in the country.

In 2012, the National Crime Records Bureau pointed out that the crime rate in Kerala (455 per lakh of population) was more than double the all-India average (196). What is more interesting is that Uttar Pradesh (96) and Bihar (127) reported much lower crime rates.

Going by data alone, it appears reasonable to conclude that God's Own Country, Kerala, is the most dangerous place to live in, while UP and Bihar are among the safest.

Going by data alone, it appears reasonable to conclude that Kerala is the most dangerous place to live in, while UP and Bihar are among the safest.

The missing element here is that these conclusions are drawn from reported crimes and not the incidence of crimes. This breathes sense into otherwise befuddling data. Being more educated and empowered, Keralites are more likely to report crimes than those in other states. The effectiveness and approachability of the police could be another reason.

To complete the story, the crime rate in Somalia (1.5), Iraq (2) and Libya (2.9) is minuscule compared to crime rates in countries such as Sweden (6456) and U.K (4447). This further illustrates that areas with high crime tend to report a low crime rate statistically.

A high crime rate may, therefore, be a reason to celebrate! Ironically, what we cannot infer from crime rate statistics is precisely what they are supposed to define -- the actual crime rate.

Evidence-based policy making: a myth?

The government collects a plethora of data to frame better policies. The idea is to undertake evidence-based policies instead of ideology-based ones.

The challenge, however, is in discerning the "evidence" used in policy making. The ideological inclinations of policymakers may prompt them to be choosy in the use of statistics.

A couple of years ago, Arvind Panagariya, who presently heads the NITI Aayog (which replaced the Planning Commission), came in for sharp criticism for being selective in his use of statistics when he presented the Kerala model as the Gujarat model in disguise.

Given the multitude of studies available -- often contradictory in nature -- it has become easier to find or even generate evidence to force-fit one's ideas...

"[Jagdish] Bhagwati and Panagariya see government restrictions everywhere, while Drèze and Sen can't take their eyes off poverty and inequality," Sir Partha Dasgupta of Cambridge University had noted in his review of the books written by these economists. His comments highlight how ideological views may inhibit holistic understanding of an issue.

It is also important to note that given the multitude of studies available -- often contradictory in nature -- it has become easier to find or even generate evidence to force-fit one's ideas and ideologies.

A 2013 study by CRISIL found Kerala to be the most equitable state in the country. Around the same time, the National Sample Survey Organization (NSSO) came up with its report which found Kerala to have the highest economic inequality in India. The only significant difference between the two studies was in their methodology and it is astonishing to observe the impact it had on the results.

A question of motivation

The fallacy could be in considering statistics to be objective and impersonal. Its subjective nature is reflected in how data is derived and interpreted. The appropriateness of the methodology of the study or survey, the quality of data cleaning, and the motivation of the individual(s) conducting the research are some of the factors that result in its subjective nature.

Perhaps the biggest challenge is not in interpretation, but in guaranteeing the accuracy of the data used for policy-making.

Think-tanks, which complement the government in policy making, are good examples. Reports published by think tanks that are run or funded by the government will often side with the stance of the policy initiated by the ruling regime. It is not surprising that Mr. Panagariya is an ardent supporter of the Gujarat model.

Likewise, it is difficult to establish the credibility of reports published by think-tanks funded by corporate and private agencies. In recent years, several reports have surfaced suggesting that oil companies fund think-tanks to generate reports of climate change denial. The ideological leanings of the organization or individual conducting the study can also affect the nature and purpose of study undertaken, and the interpretation of data.

Monopoly and the concern over accuracy

Perhaps the biggest challenge is not in interpretation, but in guaranteeing the accuracy of the data used for policy-making.

In January last year, the Indian government had introduced a new methodology to compute GDP. Several economists, including RBI governor Raghuram Rajan expressed their scepticism regarding the subsequent growth figures.

Since GDP growth has become a matter of national pride and is considered to be an indicator of good governance, one cannot completely rule out the possibility that the dubious GDP data is a deliberate attempt by the government to create an "optical illusion" of high economic growth.

Governments have a monopoly over a lot of data -- including GDP -- often for good reason. There have been instances when this monopoly power was exploited for its own vested interests --interests which were, perhaps, not shared by the public it represented. If a public body feels that it can get away with misrepresentation of facts for its own survival, it might do so.

The message is clear. One needs to realize that statistics, being a product of individuals with their own views, is a subjective tool.

The Greece debt crisis should still be fresh in memory. The country admitted to fudging its budget accounts in 2001 to secure entry into the Euro. What ensued later needs no reminder.

The message is clear. One needs to realize that statistics, being a product of individuals with their own views, is a subjective tool.

Before drawing conclusions from a "statistical finding", it is important to dig deeper and make oneself familiar with the ideology of the researcher and the methodology of the study. Ignorance of these underlying factors can have disastrous consequences and may lead to erroneous policy-making. It helps to be sceptical and cautious.

Like Us On Facebook |
Follow Us On Twitter |
Contact HuffPost India

Also see on HuffPost:

Yoga poses in French

More On This Topic