Making sense of the FY22 final GDP growth estimates

The final FY22 national account estimates shed light on dynamic trends in aggregate economic activity (GDP) and the heterogeneity within the aggregates. These are informative for assessing the movements in the real economy, the domestic counterpart of external balance and the policy arsenal that can potentially be deployed to correct non-feasible and unsustainable internal and external imbalances.

Thanks to the BBS for releasing the final estimates of GDP growth in FY22 ahead of their provisional estimates for FY23. The final estimates are slightly lower at 7.1% relative to the provisional 7.3% projected in May 2022. This is remarkable only as a break in historical regularity. The final estimates have tended to be higher than the provisional in recent years.

The final estimates capture the direction right. All the leading indicators of the state of the economic activities painted the same structural trends as depicted in the BBS estimates for FY22. The level of estimated growth itself continues to elude corroboration.

The delta of growth in FY22

The Greek symbol delta is often used by economists to represent changes in any variable. The delta of GDP growth in FY22 was 0.2 percentage point relative to FY21 – not completely unremarkable if factual.

The magnitude of the increase hides the broader base of estimated expansion. Growth in large industry jumped from 10.6% in FY21 to 15.7% in FY22 and in cottage industry from 10.3% to 11.1%. In agriculture, there were significant increases in growth in crops & horticulture and animal farming. Construction, wholesale & retail trade, transport, and real estate sectors recorded impressive increases in growth as well.

A few large pockets of distress include fishing, where growth appears to be on a secularly declining path; Small, Medium, and Micro industry, where post Covid recovery was strongest in FY21 followed by a sharp fall in FY22; and primary energy, where decline is threatening to become persistent. Growth in the two recent highfliers in services – Information and communication and Human health and social work activities – decelerated.

Demand side delta was driven by exports and capital formation. Real exports grew 29.4% in FY22 relative to 9.2% growth in FY21. A private sector driven recovery in capital formation complemented export growth. Private investment to GDP ratio increased from 23.7% in FY21 to 24.5% in FY22, though still below the 25.3% pre-pandemic level. Household consumption growth slowed from 8% in FY21 to 4.6% in FY22. The growth in government consumption declined from 6.9% to 6.2%. Pent up consumption demand may have faded while pent up investment demand unravelled with a lag.

Growth in domestic investment outpaced national saving. The total investment rate increased from 31% of GDP to 32.1% while the national savings rate decreased from 30.8% to 29.4% in FY21 and FY22 respectively. The external current account deficit naturally rose respectively from 0.9% of GDP to 3.9% despite accelerated export growth. Real import growth doubled from 15.3% to 31.2%, worsening the always negative balance between the goods and nonfactors services export and imports (resource balance).

The national account deltas appear consistent with the accounts on the direction of the economy implied by the balance of payments data. Excess of domestic demand growth over domestic production underpinned the increase in the external current account deficit. Shocks to savings and investments came from both external and internal sources. With the nominal exchange rate moving sluggishly from Tk84.81 to Tk86.3 per USD and the nominal lending rate capped at 9%, this translated into pressures on foreign exchange reserves and demand pull on inflation. The latter exacerbated the supply shocks to inflation. The rest is history still in the making.

A nontrivial bug

A dramatic 158% rise in real Statistical Discrepancies (SD) in FY22 raises eyebrows. The contribution of SD to growth increased from 0.1 percentage points in FY21 to 2.6 percentage points in FY22, pointing to the possibility that the observed rise in growth could just be a measurement error!

SDs in the measurement of GDP is an empirical challenge worldwide. In theory, SD is the sum of the measurement errors in the components of both GDP, a measure of production, and Gross Domestic Expenditures (GDE), a measure of the end use of what was produced. By convention, the SD is set equal to GDP less GDE. The source data underlying the GDP estimates are generally considered more reliable than those underlying the GDE estimates. Countries, such as the United States, who can afford extensive surveys on final expenditure, consider the final expenditure more reliable than GDP.

SDs, ranging from -0.6% of GDP to 0.5%, were small with changing signs in Bangladesh until the new revised series (2015/16 base year) was introduced last year. The SD has always been positive in the new series, which goes back only till FY16, varying between 1.5 to 1.7% until FY22 when it rose to 4%. Observed SDs in the developed and emerging market economies range between -3% to 0.4% of GDP.

With improved coverage and refinements in methodology, measurement errors can be narrowed. The high level of SD, its one sidedness, and the sharp increase in FY22 deserve deeper scrutiny. Country practices indicate that Supply-Use Tables (SUT) are the most important accounting gymnastic for reconciling SDs. SUTs are tools used to check the consistency of statistics on flows of goods and services on the principle that the total supply of each product is equal to its total uses.

The IMF is expected to assist the BBS prepare or update the SUT, hopefully sooner than later. Preparation of quarterly GDP expected to be delivered this year onwards as part of the statistical reforms directly supported under their recently approved program can also help reduce the errors.

Additional odd factors

Growth levels in manufacturing, construction and services moved in a direction noticeably opposite to the level of energy growth in FY22. Mining and quarrying grew 1.1%, down from 6.9% growth in FY21, driven by 4.7% decline in natural gas and petroleum. Electricity growth declined from 11.7% to 7.8% while gas declined by 0.6%.

These happened while growth in large scale manufacturing surged a solid 5 percentage points and cottage industry growth increased 0.9 percentage points from double digit levels. Construction growth moved up 0.6 percentage point from high single digit levels. Sectors do not exist independently within the economy but instead have contemporaneous spillovers due to linkages in the production process. One cannot help but wonder how the energy dependent sectors managed to accelerate without energy, either domestic or imported.

The direction and strength of the co-movement in the real wage and GDP growth rate are also instructive. A rise in real wages usually accompanies a period of high economic growth. However, it is not guaranteed. GDP includes wages, but also profit, interest and rent.

Real wage growth declined from a paltry 0.5% in FY21 to zero in FY22. Nominal wage growth did not keep up with inflation. According to BBS data, nominal wage growth at the low end of the labour market (low paid skilled and unskilled labour in 44 occupations in agriculture, industry, and services) was stable at 6.1% in FY21-22 while inflation increased from 5.6% in FY21 to 6.1% in FY22. Real wage grew fastest in the fastest growing industrial production sectors, but only 2.7% compared with 15.7% growth in value addition in this sector.

Labour markets should have tightened as domestic production increased and employment overseas boomed. The fact that output growth and real wage growth are so far apart and divergent across sectors smells of constraints on economic mobility, rent seeking, profiteering or too much noise in the estimated level of growth, presumably more than in the nominal wage and inflation data.

More smoke than fire?

The estimated level of growth (7.1%) is in the upper tail of the global distribution of growth in 2022. Measured GDP growth has been excessively smooth since 2015 with the exception of FY20 and the provisional estimate of FY21. This is at odds with high frequency data related to growth, including data on night light intensity and electricity consumption, and growth predicted by cross country panel data models of growth drivers analysed by the World Bank (2022).

The surge in the difference between the direct estimates of production and expenditure and the disconnect between inter sectoral growth rates as well as with real wages in FY22 suggests that the bugs suspected from data smell tests may have become endemic.

Bangladesh is surely in better economic shape than it was in 2020 as evident from the revival of bustling shopping malls and jammed roads even on weekends. But the crowd scenes do not suffice to affirm exactly how well the economy recovered in FY22. Nor do the final estimates of FY22 GDP.

Zahid Hussain is the former lead economist of World Bank Dhaka Office