Problematic statistics

The Pakistan Bureau of Statistics (PBS) has been endeavoring to improve the quality and coverage of statistics in Pakistan. These efforts include the preparation and dissemination recently of quarterly estimates by sector of the GDP of Pakistan.

This was a commitment made in the on-going Stand-by Facility with the IMF (International Monetary Fund). The expectation is that the availability of quarterly growth estimates will improve the quality of economic management in the country by facilitating faster and more informed responses.

Earlier the PBS had undertaken a major exercise of up-dating the base year for the GDP estimates by sector and by expenditure from 2005-06 to 2015-16. Consequently, the new GDP series with the latter base year have become available from 1999-2000 to 2022-23. The Consumer Price Index and the Quantum Index of Manufacturing also now have 2015-16 as the base year.

Further, recently time series have been made available of the value of outputs and intermediate inputs, both at current and constant prices of 2015-16, of each sector in the national economy. All in all, the exceptional efforts by the PBS in expanding the time series data base on Pakistan must be recognized.

However, inevitably, some of the estimates have raised doubts about their reliability. The objective of this article is to identify some of the problematic statistics and estimates.

We take up first a review of the GDP estimates with the new base year of 2015-16. There has been an 11.3% increase in the size of the GDP at factor cost in 2015-16 at current prices with the base year of 2015-16, as compared to the magnitude with the base year of 2005-06.

Much of the increase is in the informal sectors of the economy, like livestock, small-scale manufacturing, construction and transport and communications. The fundamental problem is that the proper sizing especially of informal activities has to be preceded by a nationwide census of establishments and household production units to provide adequate benchmark data for rebasing.

Such a large exercise was carried out prior to the 2005-06 rebasing but this was not undertaken before the last rebasing size. Consequently, the sectoral estimates of value-added are subjective and doubtful in some cases highlighted above.

One test of the validity of the new GDP estimates is to determine the growth rate of consumption expenditure in 2015-16. This is estimated as a residual. If the growth rate is high than this indicates that the sectoral value-added estimates may be exaggerated.

The estimated growth rate in real consumption expenditure is as high as 8.5% in 2015-16, with a GDP growth rate of less than half at 4.1%. This tends to confirm the doubtful nature of the sizing of the some of the sectors in the 2015-16 GDP rebasing exercise.

Turning now to the more recent exercise of preparation of quarterly GDP estimates, there are some major concerns here also. The third quarter, January to March, has been identified as the largest in 2022-23, while the first quarter, July to September, emerges as the smallest quarter.

The revealed difference in the GDP of these two quarters is only 2.2%. This is an extraordinarily small magnitude and apparently highlights the absence of significant seasonality in the overall economy of Pakistan.

The very big surprise is that despite two crop seasons, there is virtually no quarterly variation overall in the value-added by the agricultural sector of Pakistan. The difference between the quarterly peak and trough in the sector is an unbelievably low 1.6%. Apparently, the seasonality in crop production is, more or less, completely neutralized by the opposite seasonality in livestock activities.

This only marginal variation in quarterly GDPs in Pakistan is in contrast with the revealed relatively high seasonality of economic activities in India. According to the quarterly Indian GDP estimates for 2022-23, the difference in value-added between the quarterly peak and trough is as much as 13.7%. If anything, seasonality ought to be more in Pakistan because the share in the GDP of agriculture is higher than in India.

The next new series that have been published recently by the PBS are estimates from 1999-2000 to 2020-21 of sectoral values of output and of intermediate consumption respectively in each sector of the economy. The difference between these two magnitudes is the estimated sectoral value-added.

The ratio of output to input value has been derived for each sector from the PBS series. It is truly surprising that the input to output ratio has remained unchanged in the case of the large-scale manufacturing sector every year from 1999-2000 to 2020-21 at 73.6%.

This ratio cannot mathematically remain the same over such a long period due to technological changes, divergence in sources of inputs and varying rates of increase between domestic and international prices. Therefore, this constancy in the input to output value is inexplicable. There are similar problems in other sectors like wholesale and retail trade.

There also continue to be measurement problems with other PBS series like the rate of inflation, unemployment, etc. These have been highlighted in previous articles.

There is a need to also highlight that while the PBS has added very important new time series to the data base on the economy of Pakistan, it is falling behind in key cross-sectional surveys.

The last Household Integrated Economic Survey was as far back as 2018-19. Similarly, the last Labor Force Survey was in 2020-21 while the Social and Living Standards Measurement Survey was last undertaken in 2019-20. These are very important surveys and need to be undertaken on a priority basis to highlight the extent to which there has been a deterioration in the level of inequality, incidence of poverty, rise in the level of unemployment and decline in the coverage of basic services.

Overall, The Governing Council of the PBS must play a more active role in deciding on the work agenda and the outputs for the year of the Bureau. Also, when a new data set is produced it should be subjected to careful technical examination by independent experts by organization of a due process of interaction.

Copyright Business Recorder, 2024