The plates of this volume have mostly assumed, with the orthodoxy, a labour market in which the worker who wants a job at the going wage can simply have one — in which the wage is posted, and the worker takes it or leaves it, and that is the end of the matter. But between wanting a job and holding one lies a gap the orthodox picture skips clean over: the worker must find the job, and the finding takes time, and effort, and no small measure of luck. This plate takes up the labour market as it actually works — as a process of search, in which workers and jobs grope toward one another through a fog of imperfect information — and the two things that follow from the fog: the wage the worker fixes as his own floor, below which he will not stoop, and the net the state strings beneath him while he looks. And it will emerge that these frictions, so far from a footnote to the competitive model, are the buried foundation of much that the earlier plates described — the buyer's power, and the worker's weakness, the two of them together.
Begin with the search itself. A worker out of a job does not face a single posted wage but a scatter of them — the same labour fetching different prices at different doors, for the market is not the tidy auction of the textbook but a dispersed and frictioned thing, in which identical workers routinely earn unlike pay. He cannot knock at every door; to search is to spend, in shoe-leather and in time and in forgone earnings. So he does the only sensible thing: he sets himself a reservation wage — a threshold below which he will not accept — and he takes the first offer that clears it, and goes on searching while the offers fall short. The reservation wage weighs the bird in the hand of an offer now against the better bird that further looking might yet flush from the brush. And it rises with everything that lets the worker afford to wait — his savings, the support of his family, the patience his circumstances will bear — and sinks with everything that goads him to take what he can get today. It is the floor the worker sets beneath his own labour; and it is, this plate will argue, the very floor that the subsistence theorists and the minimum-wage law each tried to set from outside, here set at last from within, by nothing but the worker's own capacity to hold out for more.
And here the frictions repay the closest attention, for they are the hidden foundation of two of this volume's central plates. Consider what the mere cost of search does to the balance of power at the wage bargain. Because the worker cannot instantly or cheaply find another place, the employer who shaves his wage does not watch his workforce melt away by morning; a few may drift off, in time, but most remain, for to leave is to search, and search is dear. The supply of labour to a single firm, in a frictioned market, therefore slopes upward rather than lying flat — which is only to say that the employer holds the wage-setting power the eighth plate called monopsony. The frictions of this plate, that is, are the source of the buyer's power of that one: monopsony is what search frictions feel like from the employer's side of the desk. And those same frictions, viewed from the other side, are the asymmetry of the very first plate. Smith's workman could not afford to wait, while his master could; and the worker's reservation wage — how high he can pitch it, and how long he can hold it — is precisely the staying power that Smith found him to lack. The cost of search is, at one and the same time, the employer's strength and the worker's weakness: the two are not separate facts but a single friction, seen from the two ends of the bargain.
Monopsony is what search frictions feel like from the employer's side of the desk — and the worker's reservation wage is the staying power Smith found him to lack.
Onto this frictioned market the modern state strings a net: unemployment insurance, which pays the worker something while he searches. And around the net has raged one of the longest quarrels in all of economic policy. The case against generosity is clean, and it follows straight from the reservation wage. Pay the unemployed worker to look, and you raise the wage he can afford to hold out for — you lift his reservation wage — and he searches longer, and his spell of unemployment stretches. The evidence for this much is among the most robust in the whole field: the more generous the benefit, and the longer it runs, the longer the average worker stays out of work, with a tell-tale quickening of his return to employment just as the benefit is about to lapse. Generous insurance, the critic concludes, subsidises the very idleness it relieves — pays people to remain unemployed, and so manufactures a part of the unemployment it exists to cushion. It is a real effect, and an honest plate grants it at the outset: at the margin, the net does slow the worker's road back to work.
But the case against the net, clean as it is, sees only one face of a two-faced thing, and the face it misses is the one the net was built for. Unemployment is a risk — a sudden, and often a ruinous, loss of income — and it is a risk against which the worker cannot insure himself in any private market, for no company will sell a policy against a misfortune that the policyholder can partly bring about and largely conceal. That one cannot buy unemployment insurance privately is a market failure of exactly the species a later group of this table sets out to catalogue; and public unemployment insurance exists to mend it. Its benefit is the worth of that insurance — the consumption it lets the worker keep through the spell, the destitution it holds off, the home he need not sell at a loss, the children who need not go without. This benefit is large, and it is wholly invisible to the moral-hazard story, which tallies the days of unemployment as so much dead loss and never once asks what the worker is to eat during them. The net does not merely lengthen the search. It makes the search survivable.
And the deepest of the modern findings cuts deeper yet, for it shows that a great part of what the critic calls idleness is no such thing. When a worker without savings loses his job, he is not free to search for the right next one; he is driven, by the bare cupboard, to snatch the first job going, however ill it fits, simply to stanch the bleeding. Unemployment insurance, by lifting that desperation, does not chiefly bribe him to dawdle; it chiefly frees him to search as he ought to have been able to search all along — to wait for the work that fits rather than seize the work that is near. The longer spell, on this reckoning, is not a distortion the insurance introduces but a correction it supplies, undoing the haste that a want of ready money, and not any true preference for leisure, had forced upon him. The economist who took the measure of this — prising the liquidity effect apart from the moral hazard — found a great part of the response of duration to be of just this benign kind: not the worker idling on the public purse, but the worker, for once, able to afford to find the right work. The net that seems to slow him is partly the net that lets him aim.
Not the worker idling on the public purse, but the worker, for once, able to afford to find the right work.
So the net is neither the pure waste its critics see nor the pure good its champions proclaim, but a genuine trade-off, with a genuine best point somewhere between nothing and everything. Set the benefit too low, and the worker is flung back on the desperation that wrecks matches and empties cupboards; set it too high, and the moral-hazard cost mounts, the searches stretching past any gain in the quality of the match into mere subsidised delay. The right level — the problem was framed by Martin Baily and sharpened by the liquidity findings that followed — balances the insurance the net affords against the distortion it works, and turns on how far consumption falls when the job is lost, on how sharply the duration answers to the benefit, and on how keenly the worker dreads the risk. There is no one answer good for all times and places, only a balance struck differently as those magnitudes differ; but the framing is the lasting achievement — that the question was never whether to insure the worker but how much, and that the insurance and the incentive are both real, and must be weighed each against the other rather than the one wished away in favour of the other.
Return, then, to the question this volume has put to every wage — and put, with this plate, for the last of its twelve times. The reservation wage, and the net that funds it, are the worker's freedom to refuse; and the freedom to refuse is the root from which all his power in the bargain grows. The whole asymmetry of the first plate came down to this: that the worker could not afford to say no, and so was made to take what was offered, while the master could afford to wait him out. And everything that lets the worker say no — his savings, his union, his outside options, and the public net strung beneath him — restores some part of the power the asymmetry stripped away, and lifts the wage up off the floor that bare desperation would set, toward the worth that his labour can actually command. The safety net, on this reading, is not merely an alms to the unlucky but a weight laid on the worker's pan of the scale — the same scale the union loaded by numbers and the minimum wage propped by statute, here propped by the plain circumstance that a worker who can eat can afford to wait for a fair offer. And so the net, like every device in these twelve plates, comes in the end to power: to whether the worker bargains from naked need or from a position of even modest security, and to how much of the wage's settlement we are content to leave to the raw desperation of the unprotected. To insure the worker is to arm him, a little, in the contest that fixes his wage; and how far to arm him is a question that turns, as every question across these twelve plates has turned, on the same unsettled weighing — of worth against power, of security against incentive, of the one against the many. The volume's plates are, with this, all set down; but the argument they were built to hold remains, as it must, and as the table has said from its first page and will say to its last — open.