The Atlantic Home

This may seem like a dumb question. An abnormally high rate of unemployment is often treated as synonymous with a depression, as by those economists who insist that until the unemployment rate reaches 10 percent (or some other number), the economic situation cannot be described as a depression. This method of distinguishing depressions from recessions is unsound; they should be distinguished on the basis of total costs, of which unemployment is only one. But the tendency to measure a recession/depression by the unemployment rate attests to the significance attached to the rate as a measure of the gravity of the bottom of the business cycle.

But there is a puzzle: why should there be a high unemployment rate just because some shock to the economy (like a fall in household wealth because of the bursting of a housing and stock market bubble) reduces the demand for goods and services across the board, at their existing prices? Why don't prices and wages just fall and, as a result, the original demand be restored?

Suppose that as a result of an economy-wide demand shock people decide to spend less and save more because they are anxious about the future. Then the average firm will experience a reduction in the demand for its products at their existing prices. It will adjust by moving down its supply curve, which will result in its reducing both its price and its output. But this assumes that its supply curve is unchanged. Yet with the fall in demand for goods and services, demand for labor will also have fallen, and so the equilibrium wage--the wage that clears the market for labor, leaving no workers unable to find jobs who want a job--will have fallen. By reducing wages to the new equilibrium level, the firm will make a further, downward price adjustment, because its labor costs will be lower. Lower prices will increase demand for the firm's products, and in turn for labor. There will be full employment.

But that is not what is observed. Here are some reasons why it is not observed.

1. When wages fall, so do incomes, and this results in a fall of demand for goods and services and therefore in the demand for labor, which is derived from the demand for goods and services. This income effect of a decline in wages is unlikely to be fully offset by a fall in prices, because a decline in a firm's costs are rarely passed on 100 percent to customers, and anyway labor costs are only one component of a firm's costs. If labor costs are 50 percent of a firm's total costs, and it reduces these costs by cutting wages and benefits by one half, its total costs will fall by 25 percent. If half that reduction is passed on in the form of a lower price, its price will fall by only 12.5 percent. Thus, wages have fallen farther than prices, and so the price effects of a wage cut will not restore the demand for labor to its previous level.

2. When demand for a firm's products falls, it can adjust by reducing output, but it cannot do anything to reduce its fixed costs, such as debt that carries a fixed interest rate. The result may be bankruptcy, which, even if the firm's depressed price exceeds its marginal cost, may result in liquidation rather than in a successful reorganization, because of the costs, delays, and uncertainty of a reorganization in bankruptcy. Liquidation will result in termination of all the firm's employees. Moreover, a fall in demand may, by preventing a firm from achieving economies of scale, cause its marginal cost to fall below the maximum price that the market will pay for the firm's products, and then reorganization will not be an option unless there is optimism about a quick economic recovery.

3. Depression-induced reductions in price may not restore demand for goods and services to its previous level while the depression is going on. As Keynes emphasized, consumers nervous about the future may reduce their spending, in favor of increased saving. This is happening in our current depression: the personal savings rate has increased in the last year from 1 percent to almost 6 percent. If the quantity of goods and services demanded falls even though prices are lower, industry will not need as many workers.

4. Workers have a reservation wage, that is, a minimum below which they would prefer to be unemployed. By placing a floor under wage cuts, the reservation wage limits the ability of a firm to withstand a fall in demand without reducing the size of its work force. Unemployment benefits invariably rise during a depression, which increases workers' reservation wage.

5. In a deflation (that is, when prices are falling and therefore the purchasing power of a dollar or other currency unit is increasing)--and we are in a deflation, although at present a mild one--failure to reduce a nominal wage (i.e., a fixed number of dollars) amounts to an increase in the real wage. But an employer cannot persuade his employees to accept a reduction in their nominal wage on the ground that it is not a reduction in their real wage. Employees will not believe him; nor, in a depression, will anyone feel better off just because his unchanged wage buys more goods because prices are falling. Anxiety about the economic environment and (in our current depression, which has involved a large loss in personal wealth as a result of the declines in housing and stock values) a desire to rebuild personal savings will leave him feeling worse off even if he can buy goods and services at lower prices.

6. Workers whose nominal wages are cut during a depression and their real wages fall as a result will have trouble making ends meet, and this is likely to cause them to feel anxious and therefore distracted at work. Personnel officers advise against general wage cuts, pointing out that (1) the entire work force will be miserable, whereas with layoffs only those laid off will be--and they will be off the premises and so their misery will not infect the remaining work force; (2) the employees who are not laid off will work harder, lest they be the next to get the ax; (3) the employer may lose his best employees, who are likely to have good job opportunities even in a depression, (4) layoffs enable the elimination of dead wood hired when the labor market was tight so that the employer had to "make do," and (5) layoffs eliminate fixed costs, such as costs of supervision, and wage cuts do not.

7. Every firm has indispensable workers, whose wages the employer would not want to cut, for fear of losing them (this is related to point (2) in the preceding paragraph). Hence the brunt of any reduction in labor costs will be felt by workers whose wages would have to be cut drastically in order to achieve essential economies, and so it is more efficient to lay them off. 

8. From the employer's standpoint, it is probably easier to estimate the number of workers needed to satisfy the reduced demand for the employer's products than to calculate the optimal wage cut, as it cannot be sure what the workers' response will be.

Notice that these factors operate independently of unionization, which in the United States today covers only a small part of the labor force outside of public employment--and public employment is largely though not entirely insulated from the business cycle. But unionization of private firms does slant the employer's choice is favor of layoffs, because union contracts rarely limit layoffs but do limit wage and benefits reductions during the term (usually three years) of the union's collective bargaining contract with the employer.

The fact that wages do not seem to adjust much during a depression or recession has often seemed economically anomalous, as implied by the phrase "sticky wages"--a phrase that is not suggestive of efficiency. But if my analysis is correct, the tendency of the quantity rather than the price of labor to decline in response to an economy-wide fall in demand for goods and services makes perfectly good economic sense.