Skip to main content

The 2000 Census: Sampling Error

Peter Skerry

The latest installment in the continuing drama over the 2000 census is about to unfold. As of June 15, current appropriations for the Departments of Commerce, Justice, and State will expire unless Congress and the Clinton administration agree on what to do about the persistent undercounts of minorities in the census. Democrats continue to insist upon statistical sampling to adjust the count; Republicans continue to reject such remedies categorically. Unless there is some compromise, we could be facing another government shutdown. And so, after two decades, this tedious but deeply divisive technical argument may finally be bringing the government to its knees. Along the way, it could threaten the integrity of one of the oldest and most basic governmental functions.

It would be one thing if this were a battle of great significance. But the dirty little secret about the census fight is that the stakes are actually paltry. Advocates of census adjustment talk as if changing the way the U.S. government counts its population is the next great crusade on behalf of the disadvantaged—one that will expand the electoral power of minorities and send them substantially more federal funds. But adjusting the census will accomplish neither goal. Indeed, the fact that it has assumed such importance is testimony primarily to the limitations of contemporary politics: rather than confront the genuine social and political problems ofthe disadvantaged, our elected officials have come to rely on administrative mechanisms to tweak outcomes at the margin.

Let’s be clear: there is a census undercount. According to its own estimates, the Census Bureau missed 1.6 percent ofthe U.S. population in 1990. By all accounts, this figure was worse than in 1980 though better than undercounts prior to that year. Let’s also admit that there is, indeed, a “differential racial undercount,” just as critics say. In 1990, 4.4 percent of blacks were uncounted, 4.5 percent of American Indians, and 5.0 percent of Hispanics, compared with only 0.7 percent of non-Hispanic whites. Moreover, this gap has widened over the past 50 years.

But these numbers need to be put in context. First, such census data are not very reliable. The bureau’s own studies show that, when asked to identify themselves ethnically or racially, people frequently give different answers at different times. For example, among those identifying themselves as “black” in the 1990 census, only 96.2 percent so identified on a subsequent survey. For Hispanics, the figure was 92.3 percent; for Asian/Pacific Islanders, 91.g percent; and for American Indians, 65.1 percent. Such numbers indicate a limit to the accuracy we can realistically expect of census data.

As for the fiscal and political implications of the undercount, they are grossly overstated. Journalists are fond of repeating that $59 billion in federal funds is distributed annually on the basis of formulas that rely on census data. Yet, according to research sponsored by the National Academy of Sciences, only $190 million of that total (0.32 percent) would have been shifted as a result of adjusting the 1990 census.

How can this be? First, population is only one of several factors in federal grant formulas. Indeed, programs designed to help distressed communities often reduce funding when population increases. Second, to the extent that population gains do lead to grant increases, the critical factor for a given jurisdiction is not merely its absolute gain but its gain relative to other jurisdictions. Many jurisdictions could register population increases from adjustment but end up worse off than they would have been without adjustment, because other jurisdictions would experience even greater increases. Third, because such grant programs typically have funding ceilings, adjustment would result in a fixed pie divided up among more people. Per capita grants would actually decrease.

Similarly exaggerated has been the importance of the undercount when it comes to drawing legislative districts. The fact is, the partisan consequences of adjustment are impossible to predict. All participants in the debate assume that adjustment would advantage Democrats and disadvantage Republicans, on the theory that currently undercounted minorities are more likely to be Democrats. But the drawing of district lines in 50 different states is subject to myriad local political factors; quite aside from the demographic vagaries, nobody knows how each of these political scenarios would play out under adjustment. And weakened partisan loyalties would make the process less an exact science than ever.

As Tom Hofeller, current staff director of the House Subcommittee on the Census, put it in 1989: “The gerrymander overcometh all. What demographics give, legislatures can take away in the dead of the night:’ Remember, during the 1990 redistricting, Democrats controlled both legislative chambers and the executive mansions of 19 states; Republicans, only three. Today the situation is somewhat reversed: Republicans dominate 14 states, Democrats eleven. In the states under their control, Republicans need not be threatened by the additional minorities resulting from adjustment. After all, these minorities could be packed into “majority minority” districts—a tactic that has helped both minorities and Republicans in the past, at the expense of nonminority Democrats. Republicans might also mitigate the impact of additional minorities by dispersing them among several districts. Such uncertainty might not assuage the fears of House Republicans holding on to a slim majority. But this does not mean adjustment would benefit Democrats.

So why all the fuss? To begin with, the Census Bureau is an easy target. The census is a classic public good: with no well-organized constituencies to defend it, the bureau is vulnerable to all comers. It has certainly been an easy target for minorities—easier than the Bureau of Labor Statistics, whose flawed unemployment data is arguably more critical to minority interests than census undercounts. But the BLS has strong defenders, both inside and outside Congress. The Census Bureau, on the other hand, has faced lawsuits from cities as different as Los Angeles and Detroit. And, even though the cities’ claims of injury from the undercount don’t hold up to scrutiny—again, given that the likely impact of adjustment would be small anyway—their beleaguered mayors persist because adjustment allows them to seem responsive to both local boosters and to minorities.


Minority complaints about the undercount are more than opportunism, though. No one wants to be excluded from what has been called “a national ceremony”—perhaps one of the few left. The census has symbolic clout for black Americans, who intuitively trace the undercount back to the Constitution’s original stipulation that slaves be counted as three-fifths of a person. Hence Al Sharpton’s taunt at a recent demonstration: “You in the media, we know you count us as three-fifths of a man anyway.”

Advocates of adjustment also make rights claims. As New York Senator Charles Schumer has argued: “The Constitution, of course, guarantees the right of every person residing in the United States to be counted” Going one step further, advocates claim that voting rights are at stake in the census fight. To be uncounted in the census, they argue, is tantamount to being disenfranchised.

But think about it for a second. There is no “right to be counted. Quite the opposite: There is, in fact, a legal obligation to cooperate with the census. Yet here we come to one of the more disturbing results of adjustment: Legally obligated or not, why would any of us go to the trouble of filling out census forms if, through adjustment, the government offered to do it for us? Mindful that fewer Americans actually mail back their census forms each decade, experts fear that adjustment could lead to “participation meltdown. And they are right to worry.

Even more troubling is the notion that not being counted is tantamount to being disenfranchised. If individuals go uncounted, they will obviously not be included in the one-person-one-vote calculations by which district lines are drawn. Yet uncounted individuals are in no way prevented from voting—or from organizing others to vote. Conversely, those who are counted but do not vote are not enfranchised in any meaningful sense.

The case for adjustment rests on a curious notion of political power, one that reduces political muscle to sheer numbers. Obviously, numbers are important. But even more important is what people do with those numbers. In essence, adjustment advocates argue that population totals ought to translate into power independently of political effort. In an era of concern about excessively thin notions of civic obligation, this is a strikingly anemic notion of citizenship.

The debate over adjustment is less about “empowerment” than it is about the logic of the contemporary administrative state. Adjustment is the kind of administrative quick fix that our political elites would prefer to squabble over than to debate how—or whether—disadvantaged minorities can build genuine political power.

Both parties are complicitous here. Republicans are loath to seem opposed to minority empowerment and rights, especially voting rights. So their counterarguments focus on statistical and legal technicalities, particularly the constitutional requirement of an “actual enumeration.” As for Democrats, they’ve focused on the undercount because it’s one issue for which they believe they have a simple remedy. Compared with other demands that minorities make—for affirmative action, increased minimum wages, greater educational expenditures—census adjustment is relatively costless. Like redistricting, adjustment is a highly technical issue not likely to arouse much opposition among the general public. And whatever costs it might generate can be postponed for a few years, perhaps indefinitely.

Unfortunately, the particular palliative of census adjustment is a cure worse than the disease. The only group that adjustment would truly empower would be the demographers and statisticians directly involved with it. And, having sat through many sessions at the Census Bureau where these knowledgeable professionals have confused one another on the subject, I have little confidence that they would be able to explain it to the rest of us—let alone make sure it actually works. Surely there are better ways to help the disadvantaged than this. Then again, they’d probably take a lot more energy—and serious thought—than simply shutting down the government.


Get daily updates from Brookings