Plate x in print

The Insurance

search, the reservation wage, the safety net; the lowest wage a worker will accept, and what raises the floor beneath it.

A wage has a floor the worker carries into the room: the least he will take before he turns and keeps looking. It is not given to him but reckoned — and a safety net, by paying for the looking, raises it.

— the calculus of this plate, after John McCall, 1970

The Monopsony left the wage suspended between two numbers — the worth of the last hour above it, the least the worker would accept below — and handed the space between them to power. But it took the lower number as a given, a floor laid down somewhere by need or by habit, and asked only who took the room above it. The floor is not given. The least a worker will accept is a figure he works out for himself, offer by offer, mostly without ever saying so, each time he turns one down to keep looking. This plate is about that reckoning, and about the single thing that most decides how it comes out: what a worker has to fall back on while he looks.

Begin where the orthodox market is not — with a worker who cannot read the going wage off a board and simply take it. Work has to be found, and the finding takes time. Offers arrive one at a time, each a draw from a spread of wages he can only guess at; to accept one is to stop searching, and to refuse it is to spend another stretch unpaid on the chance of better. John McCall set this down in 1970 as a problem in optimal stopping, and its answer is unexpectedly clean. The worker should accept any offer above a single figure and refuse every offer beneath it. That figure is the reservation wage — the wage at which the offer in hand is worth exactly as much as another spell of looking.1

1. In McCall's problem the best policy is a single cutoff: accept any wage at or above it, refuse any below, and — strikingly — the cutoff does not change with how long the worker has already searched. This fixed figure is what the literature means by the reservation wage. McCall published the result in the Quarterly Journal of Economics in 1970.

Because the threshold is set by the worth of searching on, the things that raise it are the things that make the search more promising or less painful. A wider scatter of offers raises it: real upside makes a poor offer easy to decline. Offers that arrive faster raise it: less dead time before the next chance. A lower cost of looking raises it. And — the term that matters most for what follows — a gentler fall while he looks raises it, because the pain of refusing is whatever the unpaid week takes from him, and anything that softens that week makes saying no cheaper, and so makes the worker ask for more before he says yes. The same offer, then, can be a floor to one worker and an insult to another: the one with savings and a benefit behind him, in a market thick with openings, refuses what the man with nothing would seize — and the difference lies not in the work but in what each can afford to wait for.2

2. The reservation wage rises with the average and the spread of the offers a worker faces, and with the rate at which they arrive; it falls with the cost of search and with how heavily he discounts the future. The better-funded and the more patient the searcher, the higher the floor he can hold.
⁂

The gentlest fall a society arranges on purpose is unemployment insurance, and here the plate earns its name. A worker who keeps part of his wage when the work ends can afford to go on looking; a worker who keeps nothing must take whatever first puts bread on the table. So insurance raises the reservation wage. It is, precisely, the price of the worker's patience — the thing that lets him refuse. The net is not a charity standing outside the labor market, easing its hardest cases from without; it reaches inside the wage and lifts the floor, so that the least any worker can be brought to accept rises with the net beneath him.3

3. Studies of how long people remain unemployed find that more generous benefits, and benefits that last longer, both raise the wage workers hold out for and stretch the time they spend out of work; the rate at which people return to a job tends to jump just as their benefits are about to run out.

The same fact is the root of the long quarrel over the net. The insurance that lets a worker hold out for a fitting job also keeps him out of work longer, and that delay is a real cost — to him in wages foregone, and to whoever funds the benefit. What makes the insurance worth its cost is that a dollar is worth more to a worker with no wage coming in than to the same worker in work, so steadying his income across the two states is a real gain — one the slower search must be weighed against, not wished away. A net that replaced the whole wage forever would dissolve every reason to take a job at all. So its size cannot be a matter of kindness alone; there is a level that best weighs the security it buys against the search it discourages — enough that a worker need not sell his week in a panic, not so much that he need never sell it.4

4. The optimal benefit, in the public-finance treatment of the problem, balances the value of the steadier consumption it gives an unemployed worker against the cost of the slower return to work it encourages. Like every insurance, it must trade the protection it offers against the care it dulls — the moral hazard inherent in insuring any outcome a person can influence.

Read this way, search turns The Monopsony's own argument to the worker's side. The frictions that plate named as the root of the employer's power — that a worker cannot move costlessly, and so can be held below his worth — are the very frictions that hand him a reservation wage worth keeping, for the employer cannot replace him costlessly either, and the surplus of a match that is hard to make has two ends to it. The wage comes to rest somewhere between, and the safety net is a weight on the worker's end of the beam: by paying for the wait, it raises the least he will take, and reclaims for him some of what the inversion handed the firm.5

5. Where firms post wages rather than haggle over each hire, the same conclusion arrives from the other direction: a firm must post a wage high enough to draw workers away from further searching, so a more generous net raises the wage it is forced to offer even those it never meets across a bargaining table.

The first plate of this volume found the worker owning nothing but the week ahead, unable to wait, and so brought quickly to terms — the old asymmetry Adam Smith set at the opening of the subject. The safety net is the modern reply to it. It alters nothing about what the work is worth; it alters what the worker can refuse, and a wage is fixed as much by what can be refused as by what is produced. So the question this volume has carried — worth, or power — takes a third answer beside them: the floor of a wage is the worker's outside option, the value of being able to walk away, and that value is not handed down by nature. A society sets it, in the gentleness or the cruelty with which it lets its people fall. The flagship reads the wage as the measure of a power; this plate finds the lowest reach of that power to be a thing a public can raise, simply by choosing how far it will catch the ones who let go.