Who Accounts for Change in American Politics?

By James A. Stimson

An excerpt from Tides of Consent: How Public Opinion Shapes American Politics

When the public changes, who changes? The question, like "Who is buried in Grant's tomb?," seems to answer itself. The answer seems obvious. But it isn't. Change at the margin of 1 or 2 or 3 percent, enough to dramatically shift political outcomes, can be the product—in theory—of 1 or 2 or 3 percent of the electorate. If they moved with perfect uniformity, say, two out of every hundred Americans could account for a visible impetus for change or the difference between Democratic and Republican wins at the polls. While 98 percent stood still, 2 percent could move and produce a 2 percent change

The point here is that we don't need 50 or 80 or 100 percent of all citizens to respond to some signal in order to observe meaningful net change. And while we probably need more than 2 percent for a net 2 percent change, we don't need a lot more than that.

The Passionate

There are, first, the passionate people who care a great deal about public affairs, have strong views, and form lasting commitments to one side or the other. The ranks of the passionate include the best informed and most politically involved Americans…. While it is possible to be informed about and involved in politics while remaining neutral, few choose to do so. Those who care generally take sides. Those who care passionately generally make lifetime commitments. The defining characteristic of the passionate, then, is commitment. If politics is a cause and politicians are chosen for where they stand, then change is never necessary. The committed ideologue can—and should—be constant over a lifetime.

Readers will find the cognitive style of the passionate familiar. It is the style of most of those observed in politics and equally most of those doing the observing. They are nonetheless a small proportion of Americans.

A second group, the scorekeepers, lacks the commitment of the passionate. The scorekeepers are nonideological pragmatists who trust or distrust each side equally. They tend to see politics not as a contest of worldviews, but merely as alternate teams of possible managers of government, each contending that they can do a better job. The scorekeepers are not choosing directions in their votes; they hiring managers. Where the passionate ask of politicians, "Are their views correct?," the scorekeepers ask, "Will they do a good job?"

The Uninvolved

The final group, everybody else, lacks the interest and involvement of the others, and mostly doesn't pay attention. People who thinks politics isn't important in their lives (and they probably are right), don't pay attention and don't want to be bothered. Because they are inattentive, their views are not responsive to the happenings of politics. Failing to observe the stimulus, they do not produce the response. The movements of the uninvolved have a random character. Moved by something, they are not moved by the systematic information flow seen by their attentive counterparts.

With a lot of work we could put numbers on the three. But the categories are probably not constant over time anyhow, and so I won't bother. Whatever the true sizes, one-third, one-third, and one-third is a decent approximation. We now know enough to answer the question of who produces movement. The answer is that the middle group, the scorekeepers, accounts for virtually all change over time. While a first intuition might be that the passionate must be important movers, not moving is their defining characteristic. They never switch sides. Like players in a football match, they never take off their jerseys and sit themselves on the opposite bench. So when politics changes, they cannot be the agents of change.

The uninvolved also produce no aggregate change. Not attentive to the signals that move other citizens, they move, but not in concert. Random individual movements are self-canceling in the aggregate.

Now, to explain change at the margin we now know that we are dealing with the behaviors of only a part of the electorate, a key part that has two attributes: (1) paying enough attention to respond to the common signals of politics, yet (2) not being so involved as to be committed always to one side. They then produce all of our evidence of systematic change.

So if we ask what is typical of American politics in the aggregate, we are really asking what moves the scorekeepers. Immediately, it becomes clear that all of the evidence of "typical" citizens is irrelevant to explaining change over time. The citizen attributes that matter are those of the scorekeepers alone.

And that matters. When we believed falsely that what was typical of individual citizens would be typical of the nation, it was devastating to learn, for example, that 20 percent or so did not know which party controlled Congress, which was due credit or blame for its acts. But if all 20 percent come from the ranks of the uninvolved, then that terrible lapse of political knowledge has no consequence for the electorate's ability to reward or punish.

Political economy provides another example. We have similarly know for some time that the passionate filter their perceptions of the national economy through a partisan screen. Opponents of the president underestimate the economy's real strength and proponents overestimate it. This effect is very large. And the uninvolved often seem to be talking about another economy, so loose is their grasp of the simple facts of employment and inflation. But none of these lapses is consequential. The net perception of the economy, neither biased nor ignorant, is driven wholly by the scorekeepers. It is right on the mark.

Excerpted from Tides of Consent: How Public Opinion Shapes American Politics, by James A. Stimson. Copyright © James A. Stimson 2004. Reprinted with the permission of Cambridge University Press.