.A distinction must be drawn between two different kinds of contraction of the economy, one caused by the lockdown itself and the other, which comes after the lockdown, caused by the auto-deflation of demand to understand the macroeconomic consequence of the lockdown.
Many countries in the world had lockdowns in response to the coronavirus pandemic; India was one of them. The Indian lockdown, however, was different from that of every other country in at least two respects: first, it was far more draconian than elsewhere; and second, during the lockdown when millions of workers were thrown out of employment, the government did not come forward with any programme of universal cash assistance for the workers. The purpose of this article is to explore the macroeconomic consequences of a lockdown where no cash assistance is provided to workers thrown out of employment.
We assume a one-period lockdown, which defines our period as equal to the duration of the lockdown, and abstract from foreign trade issues by assuming a closed economy. The setting is of an economy with two classes, workers and capitalists. There is a government, and the workers “normally” (that is, in non-pandemic years) consume whatever they earn.
In the period before the lockdown, which is period t-1, let I be the level of total investment, of the government and the capitalists taken together (we ignore depreciation throughout this article by assuming for simplicity that capital lasts for ever), and A the level of autonomous consumption, again of the government and the capitalists taken together. We assume that the lockdown entails a reduction in output, but there is zero investment during the lockdown while the autonomous consumption expenditure is maintained. Because of the reduction in output, there will be larger unemployment, but the workers will maintain their consumption to the extent possible by borrowing. If the magnitude of loans they can manage to get is given by C, and the output during the lockdown by Q*t (which is independently given by the intensity of the lockdown), then it follows that
Q*t = A + w.l.Q*t + C- ∆St …(1)
where ∆S is the reduction in the level of commodity inventories, w the product wage rate, and l the labour coefficient per unit of output.
Of course, the workers are not the only ones who would be borrowing. The government sector too is likely to be in deficit, and even the capitalists may be in deficit (which they would finance by running down their reserves, that is through a reduction in their net worth), though, as I argue later, this would require a particularly severe contraction of the economy, far more than anything witnessed till now. Equation (1), which is an identity cover all cases, states that the produced surplus plus inventory decumulation equals to total autonomous expenditure plus loans to workers.
The decline in output in period t compared to period t-1 will be Qt-1 – Qt*. And since
Qt-1 must equal (A+I)/(1-wl) …(2),
the fall in output will be (A+I)/(1-wl) – Q*.
Now let us suppose that the lockdown and the pandemic are over at the end of period t, so that in period t+1 the economy can resume normal activity. Since in period t there was no investment, and output had fallen, investment in period t+1 should be lower than in period t-1, which is lower than I. Let us, however, deliberately assume that it is I. At the same time, however, since the stock of inventories has got run down in period t, these inventories have got to be replenished in period t+1. But the workers who had run up debt in period t have got to repay this debt in period t+1 (together with interest, though we ignore interest payments here), and they can only do so from their wages by curtailing consumption. Hence, output in period t+1 should be:
Qt+1 = I +A+w.l.Qt+1 – C+|∆St|,
Qt+1 = (I+A– C+|∆St| )/(1-wl) …(3)
When we compare Equation (2) with Equation (3), it turns out that whether the output in period t+1 will be higher than period t-1 will depend on whether C<|∆St|. If C<|∆St|, then output in t+1 will be higher than in t-1, but if C>|∆St|, then even after the pandemic is over, output will be lower than before. But, in such a case, investment will fall, and the economy will proceed on a downward spiral with output in period t+2 being (A+I’)/(1-wl), which is lower than (A+I)/(1-wl), because I’ would be lower than I.
What happens after the pandemic depends therefore upon whether C-|∆St| is positive or negative. The answer to this question in turn depends on whether in the absence of workers’ borrowing, and output pegged at Q* because of the lockdown, there would have been ex ante excess demand or ex ante excess supply; or, put differently, given the surplus per unit of output (that is 1-wl), at the lockdown-determined output Q*, whether the total surplus generated would have exceeded or fallen short of the level of autonomous consumption of the government and the capitalists (since we are assuming investment to be zero during the lockdown). For if it would have exceeded, then with the workers’ actual consumption propped up through borrowing C, an excess demand would have got generated, which would have been met through a reduction in inventories, but this reduction would necessarily have been less than the workers’ borrowing.
Obviously, the condition of C>|∆St| would not get satisfied only in situations where there is an extraordinarily large fall in output, but otherwise it would. With the range of falls in output that the lockdown has entailed worldwide, including in India (even after taking into account the fact that the recorded gross domestic product fall represents an underestimate), this condition, though given in a stylised form here, will get satisfied.
Some illustrative figures may be useful here, also to defend the proposition that the stark assumptions of the model do not invalidate its conclusions. Let the magnitude of surplus in the economy, which is output minus the post-tax incomes of workers (and of petty producers, though they do not enter this model), be 60% of output, and let the level of investment be 35%. By our assumption that investment is zero during the lockdown, the output has to fall by almost 60% (that is, 35 divided by 60), for any ex ante excess demand to emerge in the absence of workers’ borrowing. True, investment may not fall to zero during the lockdown as we have assumed, but even if investment does not fall to zero but to, say, 15% of the pre-lockdown output compared to the 35% that it was prior to the lockdown, even then output would have to fall by more than 33 1/3% if there has to be ex ante excess demand in the absence of workers’ borrowing. This is more than even the maximum fall in output (32%) that is suggested by knowledgeable persons to have occurred during the lockdown in India, which was the highest among all the major economies. It follows, therefore, that C>|∆St| is a condition that would be satisfied.
There is an additional reason why this would be the case. We have assumed that autonomous consumption expenditure of the capitalists, and the government is always met through adequate supplies even during the lockdown. This, however, does not really happen. In the medley of the lockdown, when some sectors are afflicted by limited supplies and others are rationed on the demand side, if for no other reason than at least for the fact that people keep to their homes and do not venture out, thus postponing their purchases, there is no reason why our assumption should be satisfied. In such a case, the demand created through workers’ borrowing is met not by a running down of inventories with the sellers, but indirectly, at least in part, by a reduction of purchases by the government and the capitalists for their autonomous consumption, which again would make C>|∆St|.
Of course, if the requirement of autonomous consumption is not met in period t, then this would create some additional (deferred) demand in period t+1. But demand deferred is, to a significant extent, demand lost. If, for instance, the purchase of some commodity is avoided in period t, then this does not mean that in period t+1 twice that amount of
the commodity would be purchased. A part of the demand from period t simply gets lost because it got deferred. The additional demand caused by such deferment therefore would not be very large. Besides, whatever additional demand is created in this manner will only offset the additional demand that would have been generated by the need to replenish inventories that is |∆St|. This is because the need to decumulate inventories in period t would have been correspondingly restrained by the fact of the postponed purchases.
Since from our previous discussion C > -∆St│D=0, where the right hand side denotes the inventory depletion if there was no deferment of purchase at all (that is, D=0), and since actual inventory depletion –∆St = -∆St│D=0 - D, and since actual deferred purchase Da< D, we can say
C > -∆St│D=0 = -∆St + D > -∆St +Da
whence it follows that the additional demand because of replenishing inventories and the actual deferred purchases will be less than C (that is, C> -∆St + Da).
It follows, therefore, that our conclusion that the output after the lockdown will be lower than before is a robust one.
The reason we have adduced for output after the lockdown being less than before is reminiscent of what Schumpeter had called “auto-deflation,” namely the reduction in demand that arises because of the need to repay debts that had been incurred in a previous period. This fall in output, in fact, will set off a downward movement in the economy. The economy will keep going down because of declining investment, since the degree of capacity utilisation would be dwindling, until it settles down at a stationary state with zero investment. Since output in this state will equal the value of the multiplier times the total autonomous consumption, there will be very substantial unemployment in the economy.
The only way this downward drift can be arrested is through an injection of demand into the economy by the government, either through an increase in the autonomous expenditure of the government (larger healthcare expenditure is an obvious candidate here), or through transfer payments to the workers. This latter course has been advocated by many in India who have suggested that the government should give Rs 7,000 per month to every family (apart from 10 kilograms of foodgrains per month to every individual). The point to note is that this is not just a relief measure; it is also a way of stimulating the economy that would otherwise be on a downward course.
Put differently, a distinction must be drawn between two different kinds of contraction of the economy, one caused by the lockdown itself, and the other, which comes after the lockdown, caused by the auto-deflation of demand. The latter, which is usually ignored, arises from the fact that the workers (including in the informal sector) get indebted during the lockdown, which they have to pay back through restraints on their level of consumption after the lockdown gets over. The proposition that, because the lockdown is over, there is no further need for providing any cash transfers to the workers is invalid; even though, formally, the end of the lockdown may mean that workers are now in a position to earn their living through employment, this employment itself would be elusive unless they are supported through such a transfer. The purpose of the present article has been to highlight this fact.
[The author is grateful to Subrata Guha and C Saratchand for their helpful comments on an earlier draft.]
Published in :https://www.epw.in/journal/2020/38/alternative-standpoint/macroeconomics-lockdown.html