Declining share of consumption in GDP

Illustration: TBS

The post-budget discourses often take off from commonly observed statistics. One such statistic was the declining share of consumption in gross domestic product, C/Y. 

A young reporter called me to get a quick reaction that she could use as a quote. I was puzzled by such demands, since GDP is an accounting construct and there may be numerous ways that C, a component in the calculation of Y, could change as a share of Y.

I texted her, mentioning that,

Y = C + I + G + X – M

In fuller form, it means calculated GDP (Y) is the sum of Private Consumption (C), Investment (I) that comes from private savings and external borrowings, Government expenditure (G) and net export (i.e., export (X) minus import (M)). 

Ideally, the focus could have been on Gross National Income (GNI) and on C/GNI, which would have to take the current account balance that would include net remittance inflow over and above the balance of trade (X-M). However, the discussion below is confined to GDP (Y).

Thus, C/Y may decline over a period if the increase in Y was due to increases in any one or more of the three components -- I, G and (X-M). That is, C/Y may decrease if,

• Investment increased without resulting in an increase in household (employment and) income.
• Government expenditure increased with no or less than proportionate increase in private (income and) consumption. Increases in debt servicing at home and abroad, as well as unabated salary increases, have increased government expenditure.
• Export increased and/or import declined.

I faced a hard time convincing the reporter that it is an accounting construct and one should unbundle the reasons within the GDP equation for a possible decline in C/Y, and 'unbound' remarks should not be entertained!

When one introduces the behaviours of consumers and savers, one may further argue that : 

• Household income = disposable income + tax paid + savings; and
• Consumption depends on disposable income once the inflationary adjustments are made. Within a systemic approach, one may consider a decline in purchasing power due to inflation as an 'inflationary tax', distinct from other forms of direct (say, income tax) or indirect (say, VAT) taxes.

Since bank deposit rates have gone up, households are likely to save (relatively) more and reduce consumption. Surely, the aggregate savings will show up in investment (in an accounting sense). 

Moreover, increased export means a decline in the availability of goods and produce for domestic consumption. Furthermore, a decline in imports, on account of shortfalls in foreign currencies, is expected to have had negative effect on local consumption.

Surely, the unbundling of the GDP equation is too long for a reporter's demand for quick reaction, and yet, it remains too short for establishing meaningful links with policies. And I was unable to satisfy the reporter's demand!

The generalised explanations of a decline in consumption's share in income would suffice if changes in past assets were taken into account. We often face crises by resorting to assets accumulated in the past, which, in negative form, include liabilities as well. That is, we may borrow (increase liability) or deplete our assets to sustain a minimum consumption level during periods of crisis. 

Unfortunately, the GDP accounting in Bangladesh is yet to formally account for changes in the important 'stock' variable, assets and liabilities. No wonder we tend to glorify the (income and consumption) flows that come with increased liability for the future generation! 

Our faith in the accounting and inferences on patterns drawn from those further dwindles once we are reminded that C (consumption) is residually calculated in the national GDP accounting!

Sajjad Zohir, Executive Director, Economic Research Group. Sketch: TBS

The author is the Executive Director of Economic Research Group (ERG). Views and interpretations may not be shared by other members of ERG.