THE FANS AND FOES of the Davis-Bacon Act seem agreed on one thing. Repeat of the 1931 act requiring payment of "prevailing wages" to workers on federal construction projects would lead to a sizeable drop in the average wages paid on such projects. To Davis-Bacon's detractors, this would be good since it would save the government millions, or perhaps billions, of dollars and probably drive down private construction wages as well. To supporters, this would be bad since it would allow fly-by-night contractors to under bid reputable firms, produce shoddy work on complex federal projects and further disrupt an unstable industry.

Both sets of claims are much exaggerated. There is plenty of evidence that Labor Department wage determinations under the act often overstate the local wages for comparable non-federal work. But there are no sound studies of how often this happens or what it adds tot he taxpayer's bill. Most studies overstate likely savings substantially by assuming that, absent Davis-Bacon, no one would do business with unions at all.

That's not likely. On large projects like dams, military installations and highways, there are usually solid advantages to dealing with unionized contractors. Jobs like these need experienced craftsmen and a relatively large and stable work force.Without Davis-Bacon the government would have to protect itself in other ways from low bids by unreliable contractors, and the result might be much the same. Repealing Davis-Bacon is thus likely to produce neither large immediate savings nor the complete disruption of the unionized construction industry. Witness the negligible impact of President Nixon's suspension of the act in 1971.

Over time, however, and especially during economic downturns, the repeal of Davis-Bacon would almost surely reduce the unions' share of federal construction. This is not something the unions will accept without a bitter fight. With relations between the administration and organized labor already far from cordial, there will be strong pressure for compromise.

There are good ways to limit Davis-Bacon "horrible cases" while preserving the act's stabilizing effect on the construction industry. Back in 1931, projects that cost less than $2,000 were exempted from coverage under Davis-Bacon. That limit is now ridiculously low.

One good change would be to restrict coverage only to those projects of sufficient size to have any measureable effect on local wages. Eliminating the collection of weekly payroll data would help cut down the irritant factor. Cutting back coverage would also give program administrators time to do a better job of determining actually prevailing local wages. Working out these accomodations will bring a fairer outcome for workers, employers and taxpayers than the polemics not dominating the debate.