What everyone believes to be the case about politics often turns out to be wrong. The conventional wisdom about Election 2002 is a good example of this enduring truth.

Everyone believes that the American people want a return to big government to protect them against terrorism and to counter corporate scandals. Even congressional Republicans seem to share that belief as they rush to protect and expand programs they promised to end in 1994.

But does the public really want more government in 2002? Fortunately, we can do more than speculate about the public mood going into this election. Political scientist James Stimson of the University of North Carolina at Chapel Hill has created a measure of the public's mood based on a large number of polling results. His data indicate that public support for liberalism and government activism has been dropping steadily since 1997. In 2002, public support for big government declined to a point not seen since the start of the Reagan revolution in 1981.

Of course, the American people do support more spending on defense, a preference that began several years before Sept. 11. But support for the war against terrorism need not translate into support for a larger welfare state.

Public opinion runs counter to big government in other ways. For a brief period after Sept. 11, the public showed more trust in the federal government. But that trust has now fallen to pre-attack levels and may even have entered a new era of decline. Voters who don't trust the federal government to do what is right most of the time are not likely to vote for candidates offering more government and higher taxes.

Have Americans become more willing to give up some civil liberties in exchange for more security? Many public opinion surveys taken during the fall of 2001 showed most Americans were willing to sacrifice some civil liberties to combat terrorism. But as the shock of the attacks faded, the respect for civil liberties and limited government returned.

By the summer of 2002, the public looked more skeptically at restrictions on civil liberties. In June, 56 percent opposed preventing terrorism at the cost of civil liberties while 40 percent supported "all necessary steps" against terrorism.

What does all this mean for the election? Stimson's research suggests the change in policy mood could benefit Republicans who are more inclined to campaign against "Big Government." Stimson believes the GOP could gain about 12 seats in the House and three or four seats in the Senate. But three caveats complicate that conclusion:

First, the Republicans may not recognize and act upon the pro-defense, anti-big-government policy mood of the nation. House Republicans seem willing to spend tax money and expand government in misguided pursuit of electoral victory.

Second, mid-term elections usually swing against the party of the sitting president. Historically the midterm election after a new president enters the White House produces gains in congressional seats for the opposing party. The party of the president usually does poorly in midterm elections (losing on average about 20 seats), perhaps nullifying the effects of the policy mood.

But we should be cautious. In the 1998 midterm election, the president's party picked up four House seats -- only the second time that happened in the 20th century. In 2000, the Democrats gained two seats despite losing the presidency. We may have entered an era where the "midterm factor" does not matter as much as it did during most of the elections of the 20th century.

Finally, a weak economy may translate into troubles for the president and his party. We have seen since July a decline in consumer confidence in the economy. So far, though, this factor has not greatly affected President Bush's approval rating.

All this indicates President Bush and the Republicans could do better than expected in November by following the public's mood in favor of reducing government and strong defense against terrorism. They could argue that less government means renewed economic growth. For the moment, however, both the president and congressional Republicans seem determined to show the electorate they can tax and spend as much as any liberal. Their misreading of the public's mood may lead to electoral defeat on Nov. 5.

John Samples is director and Patrick Basham senior fellow of the Center for Representative Government at the Cato Institute. Samples also is the editor of the new book, "James Madison and the Future of Limited Government."