For the past four years, Republican reformers have been warning of political disaster ahead. Our party's ideas have fallen behind the times, we are losing key demographic groups, and we have suffered disappointment and defeat in four of the six national elections since 1998.

You'd think that such a record would demand reform.

Yet most conservatives prefer to rationalize the dismal trends. They say: There's nothing wrong with conservatism. We lost in 2006 because of overspending and Iraq. We lost in 2008 because of the financial crisis and John McCain. There's no need to change or adapt. All we need to do is return to first principles. Rush is right!

Last week, Gallup delivered seeming confirmation to these reform-averse conservatives. The proportion of Americans who identify themselves as conservatives has jumped by three points in the past 12 months, from 37 percent to 40 percent. That rise re-establishes conservatives as the single biggest ideological bloc in the electorate. (Last year, conservatives tied moderates at 37 percent each.) Meantime, only 21 percent of Americans identify themselves as liberals. Good news!

Or is it?

Let's think harder about those numbers and what they really mean.

A big conservative bloc does not automatically translate into a big Republican vote. Republicans managed to lose elections in 1992 and 1996, and lose the popular vote in 2000, even though conservative voter identification in those years reached 43 percent, 40 percent, and 40 percent, respectively.

Obviously, 40 percent of the vote is not a winning margin in a two-party system. Even if we assume that every single one of those self-described conservatives votes Republican, the GOP still needs to pick up another 10 points or so from moderate voters. Ironically, it is the very strength of conservatism that makes it more difficult for Republicans to appeal to moderates. The Democrats know that their core group, liberals, is relatively small. Better still, the liberals know it, too. Sensing their own weakness, liberals allow their party to make compromises: to position itself closer to the center, to say things that appeal to moderate and conservative ears.

The large conservative bloc within the GOP feels no such need for compromise. It feels free to chase moderates away. Who needs them? The result is to divide the American political system between a cohesive and highly ideological Republican Party and a less ideological but larger Democratic coalition extending from the left through the middle.

Nor is it a valid assumption that all those who call themselves "conservative" will vote Republican. Voters do not always use words in the same way that politicians and pundits do. Which was the "conservative" position on the Dubai ports deal: The Wall Street Journal's or Lou Dobbs'? How do we reconcile the "conservative" view that the Social Security trust fund is an illusion and the alternative "conservative" view that the trust fund has been pillaged by unscrupulous politicians? Voters may describe themselves ideologically, but that does not necessarily mean that ideology ranks at the top, or even very high, on their list of concerns.

What does dominate their concerns is actual performance. Bill Clinton was able to win re-election in 1996—despite the conservatism of the electorate—because he could plausibly claim to have delivered prosperity, welfare reform, and reduced budget deficits. Republicans were wiped out in 2006 and 2008 because GOP policy did not translate into benefits voters could see.

Maybe we just got unlucky. But if the reformers are right about the obsolescence of Republican policy ideas, then those ideas may not work any better next time. Even conservative voters would rather pay a 40 percent marginal rate on a rising income than a 36 percent rate on a stagnant one.

A last concern: the demographic trends of the electorate. Voters under 30 look very different from voters over 30. Gallup divided the sample into four categories by age: 18–29, 30–49, 50–64 and over 65. Conservatives were the largest block in each of the latter three groups: 48 percent among the over 65s; 42 percent among the 50–64s; 41 percent among the 30–49s. Under 30, however, moderates were the largest group, with liberals and conservatives practically tied: 31 percent for liberals, 30 percent for conservatives.

Maybe these young people will grow more conservative with age. But they differ from their elders in a number of ways that will not change: They are less white, less religious, and more highly educated—all risk factors for voting Democrat. If conservatism draws its strength from those segments of the electorate that are fated to fade in the years ahead, conservatism will fade with them. Thus the urgency with which reformers make their case and seek conservative recruits: Join, or die.