The absence of any visible impact on gross domestic product (GDP) growth rates during years of political turbulence or natural disaster should in itself lead us to question our methods of calculating economic activity. It seems likely that the methods used to calculate value-addition and expenditure, while serviceable in normal times, fail adequately to account for losses in value-addition and expenditures due to disruptions caused by strikes and blockades during a political turmoil. Here are some concrete examples:
CONSTRUCTION: This sector accounted for 7.1 per cent of GDP in fiscal year (FY) 14-15. Residential and non-residential building and katcha houses are the major components. Output of residential and non-residential building construction and gross value-addition are estimated indirectly through the commodity flow method. The commodities include cement, iron and steel, bricks, timber and round wood, and fixtures and fittings. It is assumed that 31.88 per cent of the gross value of construction is value-added. It may be reasonable to expect a stable relationship between the flow of these commodities and the gross value of construction in a normal year. By contrast, in a disruptive year, the mere availability of these commodities is no guarantee that they are actually used in construction. The impact of blockades and strikes is completely missed. In addition, the method of computing value-addition in katcha house construction does not capture the impact of disruptions, because the growth rate used to extrapolate the current quantity from the 1981 benchmark is taken from surveys done several years ago. It is likely that political disruptions do not extend deep into rural Bangladesh to affect katcha house construction, but if they do the impact will not be captured.
WHOLESALE AND RETAIL TRADE: This accounted for 13.5 per cent of GDP in FY14-15. The commodity flow method is again used to estimate gross value-addition. Two fixed coefficients enter this calculation: one is the proportion of the commodities marketed, which varies from 20 (capital goods) to 100 per cent (consumer goods), and the other is the trade margin, which varies from 18.5 per cent (capital goods) to 60 per cent (vegetables). Disruptions in inter-district transportation caused by blockades and strikes reduce the marketed part and perhaps also the trade margins because of manifold increase in transport cost. But the coefficients are fixed. So the impact cannot enter the accounting.
TRANSPORT, STORAGE, AND COMMUNICATION: This accounted for 10.5 per cent of GDP in FY14 and nearly 10 per cent in FY15. Value-addition in privately organized transport is derived on the basis of net annual earnings per vehicle and the number of vehicles in operation. Net annual earnings per vehicle are assumed constant, as reported in benchmark surveys. The number of vehicles in operation presumably is calculated by adding new annual registrations to the 2005/06 benchmark stock of vehicles. A similar method is followed to calculate value-added in unorganised non-mechanised transport. That said, political turmoil likely reduces the net annual earnings per vehicle by reducing the number of days the vehicles are able to operate during a year. They also destroy a portion of the stock of vehicles. Neither of these effects is captured because net earnings are assumed unchanged at base year prices and only new registrations are considered in computing the stock of vehicles currently in operation.
PUBLIC ADMINISTRATION, EDUCATION, HEALTH, COMMUNITY AND SOCIAL SERVICES: These together accounted for 20.4--20.7 per cent of GDP in FY14-15. Gross value-added estimates are based almost entirely on compensation to employees or on assumptions of fixed income per person. Neither of these capture the severe disruptions in service delivery during political turmoil.
Taking these examples together, it appears that nearly 52 per cent of GDP accounts on the production side fail to capture the impact of political disruptions because of the way they are constructed; that is, based on measures of inputs due to data constraints. The possible exceptions are agriculture and industry, which together account for about 44 per cent of GDP.
In principle, the methods used should capture the impact in manufacturing and agriculture because survey data are used for the estimation of gross value-added (GVA) in large, medium, and small scale manufacturing for the base year 2005/06 and the Quantum Index of Industrial Production (QIIP), computed on a monthly basis, is used for extrapolation. However, there is a six-month lag in QIIP data. So for instance when Bangladesh Bureau of Statistics (BBS) prepared the GVA for large and medium scale manufacturing in FY15, it extrapolated based on data reflecting industrial performance in turmoil-free months. Also, the assumption of constant output-value-added ratios is questionable during disruptive periods when the costs of supply of raw materials and intermediate inputs rose sharply due to supply chain breakdowns.
In agriculture, the turmoil is less likely to have affected the production of goods that do not intensively use intermediate inputs, such as fertilizer and diesel. However, the waste rates and input coefficients, both of which are assumed constant, are likely to have been impacted by the turmoil, as evident from a variety of anecdotal reports. For instance, the FY15 turmoil started at the peak of winter when the aman (the second largest rice crop) had just been harvested and vegetables were ripening on the farm. With a disrupted inter-district transportation system, a lack of appropriate storage capacity and their need for cash, farmers were unable to supply aman rice and vegetables to urban markets and were forced to sell in local markets at depressed prices. Waste rates increased and farm incomes decreased. Disruption in the supply of fertilizer and diesel is likely to have increased the inputs' coefficients of boro paddy, which was being planted when the turmoil started. Input coefficients in boro are generally nearly twice as large as in aman or aus.
Thus, in practice, the adverse impact of political turmoil is unlikely to have been properly factored in the calculation of value-addition in agriculture and industry.
EXPENDITURE ESTIMATES: On the demand side, private consumption expenditures, accounting for 71.4 per cent of total expenditure in FY14, are estimated indirectly by applying growth rates of various groups of items derived from past Household Income and Expenditures (HIES) surveys to arrive at the private consumption estimates for non-HIES years. The last HIES was done in 2010. Therefore, by construction, the impact of political turmoil on private consumption expenditures is not captured. Public consumption, accounting for 5.2 per cent of GDP in FY14, is based on budget data on compensation of employees and net purchase of goods and services. Public expenditures on these inputs tend to be invariant to political turmoil.
Total investment expenditures accounted for 28.6 per cent of GDP in FY14 and nearly 29 per cent in FY15, of which construction was nearly two-thirds. Estimates of investment in construction are based on the commodity flow method, which fails to capture the disruptions in construction investments caused by strikes and blockades. Expenditures on machinery and transport equipment are estimated from the value of domestic production and imports of capital goods adjusted for distributors' margins, dealers' profit, transport and other charges. The impact is captured to the extent the estimates of domestic production and imports capture the impact of strikes and blockades.
In principle, the export and import data should capture the impact, but lags in data availability at the time of making the preliminary estimates can cause omissions. Export growth declined in the FY07 episode but rose in FY14. Import growth declined in both cases.
In sum, the failure to capture the impact appears most pronounced in private consumption and private investment expenditures.
CONCLUSION: The fact that the preliminary estimate of GDP growth in FY15 exceeds expectations based on factoring in the losses due to another prolonged political turmoil comes as little surprise. The Centre for Policy Dialogue estimated the loss at 0.55 per cent of GDP covering 60 per cent of the economy. The World Bank estimated the loss at 1.0 per cent, covering all sectors of the economy. If we net out these loss estimates from BBS's preliminary FY15 growth estimate, the growth net of political turmoil related losses is likely to have been in the range of 5.5 to 6.0 per cent; a range most analysts and observers of the Bangladesh economy would find quite reasonable as an estimate of actual GDP growth in FY15.
A proper assessment of the damage to growth done by strikes, blockades, and associated tensions and uncertainties has both political and economic value. It is politically valuable to build a public consensus against disruptive politics. It has economic value as an input in designing a policy response to the damages for which the victims' choices are not to blame. The damage is the consequence of a failure of collective action. Hence, there is expectation and demand for policy support. The damage assessment provides a basis for deciding the size and form of policy response.
[(Zahid Hussain is Lead Economist (Bangladesh) and Johannes Zutt is Country Director (Bangladesh, Bhutan and Nepal), World Bank. Feedbacks may be sent to Mehrin A. Mahbub e-mail: mmahbub@worldbank.org)]
© 2026 - All Rights with The Financial Express
