Modern economics is a data-dependent discipline. Hence, economic analysis is sensitive to the quality of the relevant data. Important economic variables influence, and in turn are influenced by, many other variables in certain manner. For example, consumer expenditure is known to be strongly influenced by income. A quantitative estimation of the sensitivity of consumption to income (as expressed by marginal and average propensity to consume) is important for business and government in preparing their business strategies or plans. Such an estimate depends crucially on data of these and other related variables. If the data are inaccurate, the estimate will differ from what the sensitivity actually is. This means the estimated value or forecast of consumption from the estimated income data will not be the same as what it will actually turn out to be.
Does this type of errors matter? Yes. It does. This is illustrated with two examples. Suppose a large business firm produces and sells a consumer good in the domestic market. The demand for the consumer good will be influenced by income. If the income estimate for the future is overstated, the forecast of sale will also be overestimated by the firm. If the firm produces according to its forecast and the actual income turns out to be lower than the forecast income when the future materialises, it will have overproduced. This will reduce its profit or increase losses.
The government can also get misdirected by incorrect forecasts. A good example is the rice price spiral early in the year 2010. The agricultural ministry had loudly claimed a massive bumper harvest of aman crop. The general expectations must have been that the rice price would fall or at least stabilise. But surprisingly, rice price started increasing squeezing the budget of the ordinary people.
Assured of an adequate supply due to the bumper harvest the government did not consider it necessary to import rice which depleted its stock. The private importers did not risk importing rice for the same reason. Thus private stock was also not replenished. When the expected surplus turned into an actual deficit due to a poor harvest, a shortage developed in the rice market. Consequently rice price started increasing. (see M. A. Taslim, 'Off the mark: It ain't so', bdnews24.com, August 20, 2010 for a fuller discussion.) This undesirable market outcome could have been prevented and import cost reduced if the rice output estimate were more accurate.
At about this time of the year several multilateral organisations such as the World Bank and International Monetary Fund (IMF) publish their forecasts of the growth rate of the gross domestic product (GDP) of the country. Their forecasts are usually less than the forecasts of the government. This has now become almost a routine that the Finance Minister (FM) angrily denounces these forecasts and assures the public that its forecast is more accurate. The great majority of the population have no clue of, nor any regard for growth. But many of the small number of people, especially economists, who do understand it mostly regard the World Bank/IMF forecasts more credible, which brings forth even more testy comments. Government statistics apparently has a credibility problem. FM's claim is not helped by what a former secretary of Statistics and Informatics Division said about how the GDP estimates are prepared by Bangladesh Bureau of Statistics (BBS). ('GDP decided first, calculations done later', Prothom Alo, 23 August 2020.)
GDP is perhaps the most important number of economics. Virtually all macroeconomic variables such as consumer spending, investment, export, import, tax revenue, interest rate, money and prices are related to it in some definite ways. For example, tax revenue rises when GDP increases and conversely. If there is a widespread discrepancy in such relationships, suspicion about data accuracy arises.
FM has forecasted that the GDP growth rate will be 7.4 per cent for FY2020-21. With the devastation wreaked by Covid-19 this rate appears to be rather high. Nonetheless, it could have been credible if the last year's growth rate were a smaller number such that this rate might have been attributed to a rebound. But BBS has already finalised the last year's growth at a robust 5.2 per cent. This has been widely perceived as an overestimate. To have a further 7.4 per cent growth in this corona-ravaged year beggars belief. The forecasts of World Bank and IMF, 1.6 and 4.4 per cent respectively, seem more likely.
The table below puts together some indicator variables that usually have a strong positive correlation with GDP. The data for these variables are generated by some government agencies independent of BBS. (Data on similar variables were utilised by Arvind Subramanian, former Chief Economic Adviser to the Government of India, to prove overestimation of Indian GDP.) What is very striking is that all the numbers indicating the growth rates of the non-monetary variables for 2019-20 are negative. The numbers for 2020-21 (first 5 or 6 months) are somewhat better with tax revenue and electricity demand becoming positive and export improving to a much smaller negative number.
A reduction in export is a reduction in export production and hence, in GDP. Nearly nine-tenths of the imports of the country is production-related, viz. capital machinery, intermediate and primary inputs. A reduction in their import necessarily implies a reduction in production or in inventories.
Tax revenues, particularly VAT, are strongly related to income and hence, GDP. Bangladesh Bank data show that the GDP growth rate was always positive during the last four decades. The growth of the tax revenue was also positive and mostly much above the GDP growth rate except for the year 2019-20 when it turned negative despite a fairly robust estimated growth. The past trends would suggest a larger reduction in GDP growth unless the tax shortfalls are blamed mainly on increased corruption or inefficiency of the tax collection authority.
A higher economic growth normally leads to a higher demand for electricity. The failure to provide electricity stifles growth. This was the logic behind the emphasis on rapid expansion of electricity generation in recent years. It increased every year during the last dozen years except for 2019/20 when the average maximum generation declined by 1.2 per cent (not shown in Table), and so did the sales revenue of PDB by 2.5 per cent. To highlight the impact of the Covid-19 pandemic the table also shows the reduction in demand in the April-June quarter of 2020 when the government had imposed a lockdown of sorts. The electricity demand sharply declined by 10.5 per cent over the same quarter of 2019. In contrast, the demand had increased by 15.8 per cent in the same quarter of 2018.
Electricity is essential for production in all sectors. The decline in demand during the last quarter of 2019-20 must have adversely affected the growth of GDP of this quarter. Thus the growth rate of GDP for the whole year could be lower than what it otherwise would have been. There is some indication of a recovery in 2020-21 with a rebound in demand.
The demand for motor vehicles of all types is also positively related to GDP. There was robust growth of 18.09 per cent in vehicle demand in the calendar year 2018 and maintained positive growth of 0.45 per cent in 2019. But the demand fell off very sharply by 24.08 per cent in the corona ravaged 2020. The registration of passenger cars declined in all three years perhaps suggesting a decline in demand due to non-income factors and a switch to other types of vehicles. But note the much higher reduction in the corona year 2020.
Monetary variables are more difficult to interpret. Normally higher growth rates of money and credit are associated with a higher growth rate of GDP, sometimes with a lag. But it is also possible that the former is dissipated in lower velocity of circulation or causes inflation rather than growth. Although the total credit and money supply increased, there was a considerable decline in private credit during both 2019-20 and 2020-21. Since the private sector accounts for 85.0 per cent of GDP, such a decline points to a decline in private investment. This has worried economists as well as the government for the last couple of years since it is the most important driver of growth. The government perhaps attempted to compensate for the lack of private investment by massively increasing its own investment. While it might have done so, it could fuel inflation in future if no offsetting measures are taken.
No single indicator variable will give much indication of the actual growth of GDP. But when all of them are taken together, as done in econometric studies, they may give an indication. The preponderance of negative growth numbers for the non-monetary variables for 2019-20 and 2020-21 does suggest low growth rates, which has led many economists to be sceptical of the ambitious forecasts of the Ministry of Finance.
If the Ministry really wants the economists to have more confidence in its estimates, it could do two things. First, it should transparently explain how it arrives at the forecasts. Second, it should help turn BBS into a demonstrably competent and autonomous institution reporting only to the president of the republic.
The author is a professor of economics, Independent University Bangladesh.