Benjamin Disraeli once defined "a practical man" as one who "practices the errors of his forefathers." In poking fun at those naysayers who tend always to deny the practicality of reasonable social change, the great spokesman of popular conservatism had a point.

Today's economy is racked by the twin evils of stagnation and inflation. So long as we persist in restricting policy options to the "practical" measures of aggregate fiscal and monetary policy, we will not be able to cure stagflation. The underlying problem is a structural flaw in the economic mechanism that needs to be corrected directly on the level of the individual firm.

Let me give an example. Suppose that wages plus benefits of the average General Motors automobile worker come to $24 per hour. This means the cost to GM of hiring one additional hour of labor is $24. If GM is trying to maximize profits, it will hire (or lay off) workers to the point where the additional revenue created by the extra hour of labor is no more nor less than the cost, in this case $24. The average revenue per hour of labor will naturally be higher--say, $36, to cover overhead, capital, profits and the like.

So far the story is rather standard. Now imagine that the United Auto Workers decide to try a somewhat unorthodox form of labor contract. Instead of having each employed member receive a wage of $24 per hour, the UAW agrees that every worker will accept as compensation a two- thirds "share" of GM's revenues per worker. At first glance, it might appear there is no difference: in both cases the employed worker is compensated $24 per hour.

But how does GM see things? Under the old contract, the company had no incentive to expand employment because the cost of an extra worker equalled the additional revenue that the worker brought in: $24 per hour. Under the new contract, if GM hires an extra worker, its total revenues go up by $24 per hour (as before) but its total labor costs now increase by only two-thirds of $24, or $16 per hour. If the company can find an extra worker to hire, it now stands to clear a profit of $8 per hour. Under the new contract, GM has an incentive to resist layoffs and, with available unemployed labor, to expand production. When production is expanded, GM automobile prices must come down because more GM cars can be sold only if their price is lowered relative to Fords, Toyotas and the rest.

Next, suppose that not only GM but all of the Fortune 500 companies are put on the new contract system at a time of unemployment. Now as each firm expands, its new workers spend their wages on the products of other firms, creating new demand for autos, increasing GM revenues and further encouraging GM to expand.

The expansion ends when everyone in the economy seeking work has a job. In each industry, the invisible hand of competition and the visible hand of collective bargaining determine compensation and employment levels just as they always have. The only difference is that now there is full employment, and labor and management are negotiating about the "sharing ratio" (two-thirds in the example) instead of the money wage ($24). The average worker, as well as the economy as a whole, is better off under a revenue- sharing system because of its built-in bias toward eliminating unemployment, expanding output and lowering prices.

This approach might be dismissed as fantasy if we didn't have living proof of its real- world viability. But we do have such an example. The Japanese economy is a lot closer than the American economy to a revenue- sharing system. In good years, many Japanese firms pay up to five months or more of compensation as a profit-sharing bonus. I am arguing that it is no coincidence that such a system goes along with low unemployment and a high level of job security. The revenue-sharing system makes Japanese firms eager to expand employment and reluctant to lay off workers.

By eliminating unemployment, the revenue-sharing system makes the typical worker better off. It is true that the tenured, high-seniority worker who already has job security may now face some variability in monetary compensation. But on average, the high-seniority worker will also earn more under revenue-sharing because full employment generally means brisk demand.

By this reckoning, the working class faces a choice. The traditional wage system offers fixed compensation for those who have work, but no guarantee of full employment and a bad deal for those without work. A sharing system offers full employment to all at a variable pay which may fluctuate somewhat with industry demand but is higher on average than the wage's system fixed compensation.

The important point is that society as a whole performs much better under a sharing system, and so does the typical working man or woman. It would be a shame if a shortsighted view hindered us from seeing the big picture and kept us from moving toward a better way of operating the economy.