GDP estimates: Between a rock and a hard place

Provisional GDP figures for 2019-20 will be announced. The official number is likely to remain in the range of negative 0.6 percent to 0.7 percent. The IMF projection is of negative 1.5 percent, and SBP's latest estimation is similar. This could be the first year since 1951-52 (GDP: negative 1.8%) to have a negative economic growth in Pakistan.

The official number would depend upon the position authorities in Islamabad are ready to assume. There are shortcomings in the official recording and these are prone to manipulation. One cannot over-emphasize the need for an independent bureau of statistics. The chief statistician's position is vacant. In some countries, this position is as important as that of a central bank governor.

Without reliable and independent data, it is hard to prepare and track any analytical economic framework. For example, if inflation is not properly recorded, monetary policy decision could be less effective. Without knowing the growth in various sectors, efficacy of any policy tool cannot be measured. Planning Commission uses guesstimates in allocating development budget without knowing the impact of existing development spending.

Back in the Ishaq Dar's days, when a few economists used to shout that reported GDP numbers are higher due to statistical gimmickries (the story goes that Dar once locked the PBS team in a room till his desired growth figure was reported). We used to argue that a good component of actual GDP growth is not reflected in reported methodology. Both arguments are valid. Now the situation seems to be opposite. The reported GDP number is likely to be better (less negative) than the ground realities.

Last GDP rebasing was done in 2005-6. Thereafter, many modern services have become part of urban lifestyle. These are not fully reflected in the official GDP computation. Now these services are also worst hit by the lockdown. Think of restaurants, marriage halls, beauty salons, gyms, fashion and other retailing, tuition centers, and other education services, OPD and other modern health services etc. Just as their growth has not been fully reflected in GDP in the last decade or so, their reported decline (due to COVID) is going to be lower than actual.

Similarly, actual manufacturing decline could be higher than what has been reported by LSM. In textile, the value addition up to cotton cloth is recorded. There is no recording of apparel and dresses manufacturing. Before COVID, textile exports volumes were growing in double digits (such as garments, bed wear, knitwear, etc); but textile growth in LSM was almost zero. Now these value-added manufacturing items are the worst hit - be it exports or local sale. But textile LSM growth recorded hit would be milder. Similarly, publishing, printing products, and media recording are not included in LSM. The exponential growth in the media during the last decade is not really reflected in the GDP.

There are also other ways to play with the recorded GDP. One way is to adjust (lower) the previous year's growth to show better performance this year. The other is to maneuver with estimation in sectors such as livestock, forestry, small scale manufacturing etc to get desirable numbers. Then, the provisional data is mostly based on 9M numbers, and life was almost normal till mid-March in Pakistan.

According to Topline calculations, official provisional recording of GDP (based on 9M reporting) will be 2.5 percent. The house expects revised (based on full year) numbers to be between 0.5 to -1 percent. However, the government may use estimation of 4Q at lower provisional numbers. What is eventually reported by the government will depend upon how estimates are recorded.

For example, LSM growth recorded during the past eight years ranged from 5.5 percent to -2 percent; but small-scale manufacturing growth ranged in a tiny band of 8.15-8.35 percent. How can this be possible? The story of livestock, fishery, forestry and slaughtering, housing services and other private services is similar. All these combined contribute around one third of GDP. Their recorded growth also moves in small bands irrespective of the high level of volatility in the remainder two-thirds of GDP.

The GDP rebasing is due. It is expected to be completed by 2020-21. Inflation is already rebased. This should be done soon along with hiring of chief statistician and a vibrant team and made independent of Q-Block's influence.