News

Data doctoring

October 02, 2018

Achyut Wagle, Kathmandupost, 2018 Oct 02,

In a report released last week, the Asian Development Bank (ADB) estimated Nepal’s Gross Domestic Product (GDP) growth rate at 4.9 and 5.5 percent for 2018 and 2019 respectively. This is almost in line with the World Bank’s (WB) growth prediction of 4.6 percent for this year. These estimates are staggeringly lower than a month-old projection made by the National Planning Commission which claimed that Nepal’s GDP would grow by 7.9 percent this fiscal year and may even touch 9.1 percent next year.     

Not only GDP, Nepal’s other economic estimates are also equally paradoxical. Our headcount poverty ratio, now claimed to have dropped close to 21 percent (down from 25.2 percent in the 2011 census) of the total population is on a par with several other countries with a far higher per capita income as well as other development indices. Our unemployment figure of about only 3 percent is also plainly ridiculous. Even our own yearly GDP growth projections have fallen short by an annual average of about 3 percent over the last decade.

Doctoring data

But these untoward experiences in statistical management and estimation have failed to restrain Nepali policymakers’ predilection to engage in purposeful and deliberate ‘data cooking’ or ‘data doctoring’, often at the behest and in the interest of powerful political masters. Such doctoring involves inflating figures like the GDP and sectoral output growth rates and downplaying the real value of variables like inflation and crime rates.

Ironically, the manipulation of mainly economic data is found to have been more rampant whenever the Finance Ministry is led by a professional economist. Moreover, the current Finance Minister Yuba Raj Khatiwada is nicknamed ‘data doctor’ in central bank circles where he spent three decades. His preference for different baselines suitable for his desired inference and ability to accordingly choose statistical tools are his well-known ‘abilities’. If viewed through this prism, his controversial economic white paper issued six months ago and recent ostentatious and simply implausible growth projections of 8 or 9 percent fall easily on the same straight line.

However, it would be unfair to blame a single minster or one particular government for such over-projections and distortions, albeit with much aggravated practice this time around. Very recently, weekly magazine Himal ran a cover story on how district agricultural offices are pressured by the line ministry to present exaggerated data on agro production, mainly paddy, and how often they fail to deflect such duress. In defence of unbelievably stark differentials in key figures, government agencies and officials often take refuge in the feeble argument that these differences were mainly due to differences in computation methods and/or extent and choice of variables. As is obvious, these phenomena can only have marginal impact, like in the case of differences seen in estimation by different international agencies, say the WB or the ADB. Differences of 3 to 4 percent in estimating the GDP alone or inflation is simply not that trivial.

There are two apparent concerns. First, Nepal, and for that matter several other developing countries, lack independent professional and academic research institutions which dare to formally contest or correct government sponsored data adventurism. No government in our history honestly showed an interest in setting up institutions like credit rating agencies and enhancing the research capability of the existing universities. Whatever is reported in the media is only indicative of the extent of the problem which, due to the obvious nature of the profession, lacks required enquiry, rigor or substantiation. Second, global experience has shown that whenever powerful governments (with the intent of totalitarianism thus brewed) fail to deliver public goods as per their own promises and people’s expectations, they more often than not resort to concealment or exaggeration of vital statistics according to their convenience.

Regimes with one-party dictatorship with lesser transparency like China, Cuba or Venezuela are often blamed for both data doctoring (customising the available data according to the political vested interest) and data cooking (creating new ‘dishes’ to feed their publicity machines). Nepal too, with its very powerful government with a two-thirds majority in Parliament but failing to deliver and increasingly being intolerant of criticism, seems to be gradually falling into this very doctoring-and-cooking trap.

Continuing politically motivated data manipulation practices, at first, is suicidal to the government as it seldom serves its own political propose in the long run. Even in the short run, it jeopardises the country’s credibility which, in turn, will adversely impact the overall investment climate and foreign direct investment and productivity of the economy, in particular.

Such state-sponsored chicanery has very serious implications on the country’s economic and development policy formulation. The trend of exaggeration envelopes the realistic need of policy calibration based on factual evidence. The practice itself is antithetic to the evolving global paradigm of evidenced-based policy planning. For example, when we misrepresent reality by claiming that Nepal’s unemployment rate is only 3 percent contrary to the fact that every year about 700,000 youths enter the labour market and the national economy barely creates 50,000 odd jobs annually, it will essentially compromise the scope of formulating policy which can better take the problem head on.

Question of integrity

Above credibility and practical issues, it also is a question of integrity surrounding data collection, use and analyses, empirical research ethics and their logical and inferential exposition. From a structural perspective, the autonomy of major data handling institutions like the Central Bureau of Statistics or the Agricultural Research Council is constrained due to the very fact of being a regular public entity which can’t risk acting against the will of the government.

The solution to this chronic problem will not be resolved without creating at least one credible national-level institution for economic research that can operate completely independent of government interference, but with unconditional and enough public endowment. The existing public institutions must also be given professional independence. The best role for them would be to serve as trustworthy data banks to autonomous and able research institutions. Here again, the political will to set up such a professional research organisation seems lacking even though budget speeches have unfailingly mentioned such an imperative year after year.



Source: http://kathmandupost.ekantipur.com/news/2018-10-02/data-doctoring.html