Hours of Work: Moving Beyond Gridlock

 One not only makes sacrifices for what one loves, but also loves that for which one has made sacrifices.

Georg Simmel
Betting on a fixed roulette wheel is an example of what is known as a "sure-thing" in the field of decision theory. But why would anybody ever bet against a sure thing? In his article, "Risk, ambiguity and the Savage axioms," Daniel Ellsberg showed that decision makers consistently bet against their judgement of the best odds in situations where the better outcome is also more ambiguous.

Collective bargaining provides perhaps the best illustration of a situation where two sides are dealing with different degrees of ambiguity and therefore are inclined to place different stress on the odds. The employer can focus on the straight-forward cost calculations. But the union must always deal directly with the much more ambiguous question of acceptance by the membership of any negotiated settlement.[5] From the union side, the hourly wage rate and itemized fringe benefits are less ambiguous than the employer's calculation of cost per hour actually worked. Unions are understandably reluctant to tamper with such apparent certainty, even though the probable outcome can be shown to be unfavourable.

After I had become confident with my cost calculations, I spoke with Jef Keighley a national rep with the Canadian Auto Workers Union (CAW). The CAW is arguably among the more progressive trade unions on the issue of work time. We talked about my suggestion that unions may benefit from a move from fixed-cost fringe benefits to benefits tied to either total earnings or hours worked. I explained to Keighley that such a move could allow for the voluntary negotiation of shorter work weeks, where wanted. Here is his objection: if the employer makes a variable, rather than a fixed-cost contribution to benefit plans, any reduction in work hours would result in either a partial loss of benefits or a loss of pay, if part of the wage were used to make up for the reduced employer contribution.

What Keighley is saying is that the union can't ask workers to risk losing some pay or benefits now in exchange for the more ambiguous promise that an effective limit on work time could, in the long run, enhance the union's bargaining position. What unions would prefer to do is guarantee an existing package of wages and benefits -- which are exalted in the union's self-portrait as historical gains -- and then seek work time reductions as an extension or supplement to those gains. The weakness of such a approach might be clearer if the given level of wages and benefits were seen as the result of "trade-offs" rather than as absolute "gains".

The union defense of achieved levels of pay and benefits is understandable, but it is not entirely justified. It neglects to acknowledge that current levels and structures of pay and benefits have been achieved partly in return for concessions on the limitation of work time. Those who have lost the most from these concessions are those who are not fortunate enough to have the security of a union job. But increasingly, over the past two decades union members themselves have been suffering the consequences of a bad strategic choice.

Contrast the modern attitude of putting pay and benefits first with the words of an English factory inspector, R.J. Saunders, published in 1848 in the Reports of the Inspectors of Factories: Further steps towards a reformation of society can never be carried out with any hope of success, unless the hours of labour be limited, and the prescribed limit strictly enforced.