Crude estimations of SMEs survival

Yesterday’s headlines echoed the plight and groan of SMEs; it quoted Asad Umar who said that up to 1 million SMEs could be shut down due to Covid lockdown. There is no denying the need to support the SMEs. When textile and retail giants could ask workers to resign on their own because their survival is at stake, then SMEs are surely worse off. What’s intriguing though where does that number come from?

The number quoted by Asad seems to have come from a study conducted by Sustainable Development Policy Institute (SDPI). Its recent study titled ‘Projected impact of lockdown on SMEs in Pakistan’ uses two surveys conducted in China to assess the impact of lockdown on Chinese SMEs. Based on those Chinese surveys, the SDPI has constructed two scenarios for Pakistan.

In scenario 1: Pakistani SMEs mirror the impact of lockdown on Chinese SMEs where 20 percent of Chinese SMEs had reported that they would be unable to survive beyond one month of lockdown, and 64 percent reported not being able to survive beyond three months. In absence of a local survey, the SDPI use the Chinese survey as a proxy. It applied these percentages to the total number of SMEs in Pakistan (3.8 million) to arrive at an estimate that 0.76 million Pakistani SMEs (or 20% of total) will not be able survive beyond one month of lockdown, and 2.48 million (or 64% of total SMEs) will be unable to survive beyond three months.

In scenario 2, which is SDPI’s higher risk scenario, the report’s authors applied higher risks, assumed to be a quarter over and above the response from Chinese survey in each category. The result: about 0.95 million Pakistani SMEs (or 25% of total) are estimated to be unable to survive beyond one month, and 3.04 million SMEs (or 80% of total) are not expected to survive beyond three months. Hence, Asad Umar’s statement that 1 million SMEs could be shut down due to Covid lockdown.

But here comes the interesting part. About the same time the SDPI released its report, SMEDA also released its report titled ‘Survey report: Impact of Covid-19 on SMEs’. Their survey, based on 920 respondents (SMEs) with estimated annual sales of Rs29 billion, and an estimated total of about 19,600 workforce employed, suggests far more worrying statistics than that estimated by the SDPI.

SMEDA reports that about 59 percent of its respondents (SMEs) said they will be not able to survive after one month of lockdown, and about 90 percent said they won’t be able to survive after three months. If SMEDA’s estimates are any guide, then 2.2 million SMEs (59% of total 3.8 million) won’t be able to survive beyond one month of lockdown. This begs the question which of the two numbers should Asad Umar rely on – SDPI’s or SMEDAs.

As it turns out economics is not just a dismal science. It’s also poor science, if it can be called science at all. It relies on surveys, estimates, proxies, and other kinds of ‘jugaar’, which by the way means the same as ‘mech’ (the root word of mechanical) i.e. a clever trick to solve difficult problems.

The extent of glorified ‘jugaar’ can be gauged by the fact that the whole edifice of modern economy is based on estimations and surveys, such as GDP, inflation, unemployment. Yet there is always some method to madness.

Granted China is an all-weather friend, but deeper than oceans and higher than Himalayas friendship should not be a substitute for robust research methods. More so, when the government’s own research capacity is questionable, and it often relies on outside research for decision making in times as fluid as these.

Principles of transparency, accountability, reliability, consistency and broad values and virtues that dictate economic estimations demand some method to arrest madness. While SMEDA’s report has respondents from more than 65 cities across the country, it’s not a statistically robust nationally representative sample, which doesn’t make it a reliable survey. Likewise, SDPI’s estimations are based on very crude proxies that can’t stand scientific rigour.

This is not to make a mountain out of a molehill, nor to belittle the efforts by these organisations. The SMEs surely need government assistance, regardless of the differences in the estimated impact of lockdown. But the degree and duration of assistance will have to depend on the degree and duration of impact of lockdown. Which is why sooner not later the government would do well to demand robustness into the analysis and information basis feeding into their decisions.

As an addition to their many useful recommendations to support the SMEs, the SDPI and SMEDA should stress on the need to improve national statistics on SMEs. As it is, the total number of SMEs, their contribution to GDP, exports, and employment are mythical numbers whose last crude estimations (which were not even peer reviewed) were made in the year 2000. Since then those numbers have been blindly cited by all and sundry. (See BR Research’s ‘Is SME finance really at historic high?,’ Feb 22, 2019)

Granted that the government does not have time for a full-blown SME survey given the urgency of decisions need to be taken by the government. But one expects the government to improve its approach in the months to come; after all the PM himself has publicly stated that he expects the pandemic to last up to 12 months. Ergo, government endowments are here to stay indefinitely.

One way to go about is to task SMEDA to do a nationally representative survey every month to periodically gauge the pulse of SMEs. Another way is to kickstart a documentation drive with easy procedures and link endowments to SMEs to registration with government databases, Nadra, FBR etc.

The latter has been advocated earlier in this space, and BR Research is informed that the same advice has been given to Finance Ministry by Pakistan Institute of Development Economics. Of course, the ideal would be to speed towards unified business registry while ramping up federal and provincial statistics for overall statistics, SMEs or non-SMEs. (See also ‘Covid-economy: Big Data should be Pakistan’s New Deal’, Apr 16, 2020)