Over the last two months people who don’t know very much about statistics and econometrics have been sniping away at each other. This is not new but it’s the first time that dirty linen is being washed in public.

The world is familiar with the egos of economists. What it’s not familiar with is the egos of statisticians and econometricians.

Over my last 43 years as an economic journalist and occasional member of think tanks, I have frequently witnessed a discussion getting derailed because two obstinate data types were quarrelling over some arcane point that made no difference at all to the main matter at hand. Statistical accuracy is important but when it becomes a theological debate — is electricity fire? — we are in trouble.

I fear this is what has happened in our statistical system. Too many cooks are spoiling the broth and no one gets anything to eat. Likewise, we aren’t getting the statistics we need.

Either our data is old (like inflation weights, consumption, population, poverty, health, etc) or, even worse, it isn’t there at all. This is specially true of data that’s supposed to be collected, analysed and disseminated by the States.

The short point is that our policymakers are making policy on old and/or non-existent data. It’s like a pilot flying partially blindfolded through a heavy cloud. It’s also worth noting that data is much more accessible with the RBI or the Economic Survey. On the Ministry of Statistics and Programme Implementation (MoSPI) website it’s hard to find, if it’s there at all. Amazing, is it not, that this should be so? But it is.

Data blindness

Why are we in this mess? One reason as noted above is the proclivity to lose sight of what’s essential and what’s purely of academic value. Another is that within the government, the MoSPI is given no importance. It’s a crucial ministry which should have priority but even the Ministry of Urban Development gets preference.

It’s vitally important that the statistics ministry be involved in the design of policies and projects of other ministries so that meaningful numbers can be collected. But this is never done. Result: the statisticians have to work with poor data.

A third reason is the biases of the staff itself. There are clear political divisions and differences that hold things up. National statistics shouldn’t become an ideological wrangle.

A fourth reason is the persistent battle between outside excellence to head the organisation and internal career objectives. Much of the ministry’s secretary’s time is taken up in resolving silly arguments over administrative issues.

Fifth, there are all the technical reasons of data interpretation over which the ministry spends inordinate amounts of time without a resolution. In short, simply nothing happens.

Sixth, and this is perhaps the root cause of why MoSPI is dysfunctional. We as people have very little respect and regard for data. We prefer policies based on political preferences rather than what the numbers suggest should be done.

So see what the ministry itself says: “Many of these laws are outdated and may require revision. The Government of India had set up in 1998 a Commission to review such laws, but it felt handicapped by the non-availability of rules, and regulations related to this.”

In other words, no one gives a damn. To see how and why go here: https://mospi.gov.in/148-legal-provisions-he-statistical-system

The soft parts

The mystery is why the Modi government hasn’t tried to fix these problems as earnestly as it has tried to fix other problems. There are three relatively simple ways of doing it. There could be many more, of course.

First, involve the ministry more. Maybe the statistics minister should be the prime minister himself with a technically competent minister of state to assist him or her.

Second, develop a better cadre management policy because at the end of the day the statisticians also live in a bureaucratic environment. A poor cadre management policy is what ruined the Indian Economic Service. The RBI’s economic research department provides a good example to follow.

And third, depoliticise data by ensuring that ignorant people aren’t citing dodgy numbers. Freedom of speech need not include the freedom to treat data so cavalierly. Statistics is as specialised as aeronautics is.

There is of course a fourth way, the Chinese way. China announced last Tuesday that they would be discontinuing their unemployment data. This is the latest in a string of similar decisions. We could do that, too, I suppose. Perhaps we already have.