Aselection day approaches, fixing responsibility for the budget deadlock gets more and more attention. How will voters apportion blame for this debacle? How will their appraisals affect the results?

Pollsters are producing running tallies on the first question. Midterm elections are a well-plowed research field, hence the answer to the second question is fairly clear: This is a game where the president runs all the risks. The more voters who think he is at fault, the poorer showing the Republican ticket will make tomorrow. On the other hand, the electoral consequences of blaming Congress will be close to nil.

This asymmetry is easily explained. As an institution, Congress is not notably popular. In the five elections of the 1980s, the University of Michigan's authoritative National Election Studies found that barely half of the public with an opinion on the subject approved the way Congress was doing its job. (The president's approval ratings were a good deal higher.) But "Congress" does not run for reelection; its individual members do, and they are much more popular than it is. The same surveys found that 88 percent thought favorably about their own representative's performance. And when the votes were counted in the past decade's House elections, 94 percent of all incumbents were returned to office.

The public's distinction between Congress and its members is encouraged by those members. The distinguished political scientist Richard Fenno, who accompanied 18 very diverse representatives as they visited their districts, reported that every one of them attacked the legislature in which they served and lost few opportunities to differentiate themselves from it, Fenno concluded that "Members of Congress run for Congress by running against Congress."

No matter how unpopular Congress is, individual members suffer very little as a consequence, except perhaps in the public clamor against increases in their pay. Variations in the institution's popularity have little effect on the public's approval of their representative or the electoral success of incumbents running for reelection.

The same is not true for the president. His popularity is a precious resource for himself and his party. In a presidential election year the relationship is obvious. Popular presidents like Ronald Reagan or Richard Nixon are reelected. Unpopular presidents, like Carter, are unlikely to be returned to the White House. And if the public judges a president's performance harshly, its verdict is transferred to his successor nominee. Thus Hubert Humphrey paid the price in 1968 for the American public's low opinion of Lyndon Johnson's performance in the White House. By the same token, George Bush was the beneficiary of Reagan's rise in public esteem during the fall of 1988.

The president's popularity or lack of it has a similar effect on his party's candidates for the House of Representatives. Research covering the past 11 elections shows that every two points the president's popularity drops in his second year in office costs his party about one House seat in the ensuing midterm election. In short, the president's popularity matters; Congress's popularity does not.

The explanation for this relationship seems to be that some voters are moved by their assessments of the two parties' ability to satisfy them on the most important public concerns. The image of the president's party is largely set by his apparent success in office. As president Bush's approval dwindles, so do the prospects of victory for his fellow Republicans.

A few members of Congress conceivably may lose because of their own votes on the various budget plans considered on Capitol Hill. But they will not suffer tomorrow because the institution in which they serve suffers a decline in public esteem. Jokes on the Johnny Carson show may hurt their feelings, but not their careers.

The writers are political science professors. Raymond E. Wolfinger teaches at the University of California, Berkeley, and Richard A. Brody at Stanford University.