GDP: More data would help

The provisional GDP growth for FY20 is recorded at -0.38 percent. The SBP was expecting 3 percent growth (prior to COVID). The IMF forecast for post COVID growth was a negative 1.5 percent (pre-COVID: 2.4%). The released nominal GDP number at Rs41,727 billion (USD 265 bn) is very close the latest IMF projections of Rs41,939 billion. FY19 real GDP growth is revised down from 3.3 percent to 1.9 percent. This was bound to happen (irrespective of COVID). Had FY19 number not been revised down, FY20 real GDP growth decline would have been at 1.7 percent.

The usual methodology is that the provisional GDP number are announced based on 6-9-month actual data availability. For remaining 3-6 month, 6-9 -month growth and seasonality is adjusted to give a provisional number for full year. This year the National Accounts Committee took an exception and 6-9-month actual data is annualized by incorporating the impact of COVID in the last quarter.

Some say that such low growth rate is not possible due to 1.5 month of partial lockdown. The other group says that actual GDP decline could be higher. One cannot say with certainty without having detailed data. The good omen is that PBS is not pressurized on overstating the data (a usual practice in the past). The PBS has started to flex its muscles (though constrained by limited quality human resource). One must appreciate the less reliance of the institution on the MoF. A similar example is of independence of SBP is formulation and implementing the exchange rate policy. The PBS is an equally important institution and without correct data, policy makers are shooting in the dark.

The SBP was forecasting FY20 GDP growth at 3 percent in Jan-20 on the previous year’s provisional base. Since that base is revised down, the growth could have been higher. Some may ask why the GDP in FY19 is revised down. Well, that was bound to happen. The numbers in the last quarter last year were worse than what were earlier expected. Thus, GDP had to come down. GDP numbers take three years to finalize. It is a normal practice and PBS is being consistent.

Anyways, the revised real GDP at constant cost (FY06) in FY19 came down from Rs12,750 billion to Rs12,580 billion. The growth declined from 3.3 percent to 1.9 percent. The low base now gives higher room for growth in FY20. The FY20 real GDP at constant cost is estimated at Rs12,.532 billion. There would have been a decline of 1.71 percent based on old FY19 numbers and is -0.38 percent on revised numbers.

In FY20, agriculture is provisionally growing at 2.67 percent, industrial sector down by 2.6 percent and services by 0.59 percent.  Agriculture crops are largely insulated to COVID; but livestock isn’t. The growth in livestock at 2.6 percent could be on the higher side. But these numbers are not substantiated with actual data. Even PBS is constrained in its calculations. For details read “GDP: If wishes were livestock”.

The industrial growth is down by 2.6 percent. The sector is adversely affected by the lockdown and still it’s not clear how much demand picks up in June. LSM is down by 7.8 percent in 9MFY20. LSM methodology is not robust and is based on estimates. Not many value added segments are part of LSM – such as value added textile. And many of these are badly hit due to COVID. Thus, industrial sector could be lower if the actual data is used.

The issue could be in services sector decline of 0.59 percent. This contributes to around three fifth of the GDP. The decline is plausible seeing the real economic decline in numerous urban services due to COVID. But not all those are part of GDP computation based on 2006 constant prices. Since these services are not fully recorded in GDP, their decline should be limited too in recorded numbers.

The wholesale and retail sector is down by 3.4 percent. The contribution of this sector in GDP is almost equal to agriculture. This sector’s decline is a little surprising as a good chunk of it emanates from agriculture which is not doing that bad. Plus, the essential services kept on functioning in the lockdown. But nothing can be said with certainty without actual data.

Finance and insurance sector increase is limited to a mere 0.8 percent (previous 5-year average growth at 7.2%). That number could be contested as most of the financial services are open in the lockdown. The general government services are up by 4 percent (last five-year average growth: 8.1%). With the kind of additional government spending in the last quarter, the decline of the growth from five-year average is hard to digest.

The bottom line is that the provisional growth number has to be taken with (more than) a pinch of salt. Times are fluid and data takes time to come. The chances are that FY20 revised numbers (next year) could vary from significant provisional estimates. The pressing need is to develop capacity at PBS to expedite the GDP rebasing exercise (like it did for inflation) and it should start computing quarterly GDP data.