CONGRESS is about to pump at least $4 billion more into the public works grants so beloved by local officials, the unions and the building industry. This is an expensive, inefficient way to cut general unemployment, and the spending could stimulate the economy at the wrong time. But carping along those lines is so much wasted breath. With 22,000 local projects vying for funds, a big bill is surely going to be passed.

The most to be hoped for, then, is serious attention to the worst problems in the original $2-billion public works program enacted over President Ford's veto last year. That law was poorly drafted and gave the Economic Development Administration too much latitude and too little time to parcel out the aid. As a result, many big cities and other areas with severe unemployment got no funds, while localities with lesser problems walked off with generous grants.

This time around, the House Public Works Committee has eliminated two major quirks in last year's law. The most glaring one, included to marshal enough votes to override Mr. Ford's veto, allotted 30 per cent of the money to areas where unemployment was less serious. Another troublesome clause enebled some suburbs and tiny towns to get excessive grants by capitalizing on high unemployment nearby.

Beyond that, though, the search for equity leads into a political bramble bush. How do you choose between a sewer system for a town with 14 per cent unemployment, and 2,800 people out of work - and a detention center for a city where 8 per cent unemployment means 60,000 people without jobs? If you're a congressman, what's fair depends a lot on where you're from. For Congress as a whole, the easiest answer is to try to fund everything. If that's impossible - and even $4 billion may accommodate only one-fifth of the projects now piled up at EDA - the usual course is to legislate general criteria and leave the tough decisions to an agency and its computers, which can always be blamed when one's constituents lose out.

Thus the House committee, after hearing hundreds of conflicting suggestions, shied away from fiddling too much with the existing formulas, which tend to favor smaller towns and states where unemployment rates are high. In the Senate, though, the arithmetic is different. The Public Works Committee there has tentatively approved an amendment by freshman Sen. H. J. Heinz III (R-Pa.) that would "simplify" the allocation among the states and (by sheer coincidence, no doubt) increase Pennsylvania's allotment 28 per cent. Much more important, Sen. Heinz' approach would help 26 states in all, and thus might win a Senate majority and bring on a conference fight.

So the calculating goes on. At least Congress is learning more about the intricacies of passing out pork by computers. But the tendency to build a little something everywhere is just as strong as ever. Of course, public works bills do bring out the most parochial impulses in Congress; so this may not be an ideal test of the legislators' willingness to concentrate federal aid where the problems are most acute. Even so, the evidence so far suggests that programs calling for much tighter "targeting," such as community development, may face hard going on Capitol Hill.