Testimony Before the Subcommittee on the Census, Committee on Government Reform and Oversight, Hearings on the Census

Good afternoon, my name is Marlo Lewis. As Vice President for Policy and Coalitions of the Competitive Enterprise Institute, I welcome your invitation to discuss the Census Long Form. CEI is a public interest group established in 1984 with a current staff of 35 and an annual budget of about $2.5 million. Located in Washington, D.C., CEI works to educate and inform policy makers, journalists, and other opinion leaders on private voluntary and free market alternatives to government programs and regulation. CEI also engages in public interest litigation to protect property rights and economic liberty. CEI is supported by the voluntary contributions of foundations, corporations and individuals. We accept no grants from any government agency, nor do we accept grants from any other party that would compromise the principled positions we espouse.

Mr. Chairman, I commend you for holding this hearing. The Census Long Form is a federal action that will directly touch one-sixth of the households of this country. The information collected by the Long Form facilitates and shapes scores of federal programs. The debate over the Long Form is part of a broader debate about the proper scope and limits of federal power. It is with an eye to that wider debate that I will frame my remarks.

I.  Dress Rehearsal Bombs

According to a recent Washington Times article, a "dress rehearsal" for the year 2000 census reveals that "nearly half the population doesn’t want to be counted." In 12 South Carolina counties 46% of those who received questionnaires did not respond. In Sacramento, California, 47% did not mail back their census forms. On the Menominee Indian reservation in northern Wisconsin, 60% failed to comply. These results were a big disappointment to the Census Bureau, which took unusual steps to encourage a more robust response. As the Times article reports:

It [the Bureau] undertook a major advertising campaign in each test area. It mailed or delivered 460,444 meticulously redesigned census forms in multiple languages and placed additional forms at community centers. It generated news stories about the test and arranged for spreading the word in churches and schools and on local cable TV. It did another thing. In a test conducted as part of the 1990 census, the bureau confirmed that the mail response rate was higher when questionnaire envelopes contained a bold reminder that the law requires people to respond. So it used such envelopes in the rehearsal.

Actually, people’s willingness to fill out and mail back census questionnaires has been dropping for three decades. The rate of return was 78% in 1970, 75% in 1980, and 66% in 1990. If the dress rehearsal is any indication, almost half will not comply in the year 2000. Part of the explanation for this trend, Census officials say, is that people don’t trust the government. Many believe the information will not be kept confidential or may later be used against them.

Who can blame them? The IRS also is supposed to protect the confidentiality of the information it collects, and it’s supposed to deal fairly with taxpayers. Has it always done so? Has it scrupulously abstained from acts of intimidation and persecution? As our government has grown, it has become more intrusive, less accountable, and less respectful of individual liberty. The Long Form itself undoubtedly contributes to the growing distrust of government in general and the deccenial census in particular. In the public imagination, the census is increasingly identified with the Long Form and its impertinent questions. Common sense tells us to be wary of those who pry uninvited into our private business. Indeed, why would anyone nose around in my personal affairs if not to find something he can later use to my detriment?

Article I, Sec. 2 of the Constitution mandates a decennial census or "enumeration" for a very specific and limited purpose – to apportion congressional representatives among the states. The Census Bureau is failing to meet its constitutional responsibility to produce an accurate head count. The Long Form is very likely driving down response rates for the census as a whole. The government’s attempt to collect information beyond its constitutional mandate is interfering with its ability to fulfill that mandate. That this is happening should not surprise us. Government doesn’t do many things well, and the more functions it assumes beyond its core responsibilities, the less well it does anything at all.

Mr. Chairman, I will now present three reasons for abolishing the Long Form. I do not recommend immediate abolition; too many private and governmental interests have made plans with the expectation that Long Form data will be collected in the year 2000. But by the year 2010, the Census Bureau should return to its original mission of enumerating the population to apportion seats in the House of Representatives. After arguing for abolition, I will then respond to some of the arguments made by defenders of the Long Form.


II. Why the Long Form Should Be Abolished

First, the Long Form departs from the original intent of the Constitution. The Constitution requires a census or "enumeration" of citizens (Art. I, Sec. 2). Calling the Long Form a "census" is an abuse of terminology. It is deceptive. Ask yourselves this question. Would Congress dare to require Americans to submit information about their age, ancestry, disability, education, gender, grandparents, ethnicity, occupation, language spoken at home, marital status, race, telephone number, bedrooms, kitchen facilities, plumbing facilities, condominium fees, and the like in a survey separate and apart from the decennial census? I think not. Because then it would be clear to everyone that the federal government was demanding more information from American citizens than it had constitutional warrant to collect.

One can certainly admire the cleverness of those who decided to time this intrusive exercise so that it coincides with the decennial census, just as one can admire their shrewdness in calling it the Census Long Form. But calling the Long Form the "Census" does not make it so. The Framers provided for a decennial head count and nothing more. I can only speculate about their reasons. But they surely knew that knowledge is power. Limited government must have limited knowledge about the personal characteristics of its citizens if it is to remain limited and not try to run their lives.

Second, the Long Form is intrusive – a violation of personal privacy. If some stranger on the street – indeed, if a prospective employer at a job interview – asked for information about your ancestry, age, income, marital status, race, Hispanic origin, bedrooms, plumbing facilities, and so on, you might be inclined to tell him where to go. Just because the person collecting or processing such data is a federal bureaucrat does not make the process any less intrusive. Indeed, you can walk away from unwanted solicitations on the street, and you can tell the impertinent job interviewer that you’ll see him in court. But citizens have no legal recourse when the government duns them for facts about their private lives.

That’s what the late William F. Rickenbacker found when he mounted a legal challenge to the Long Form. Writing in National Review a few months prior to the litigation, Rickenbacker declared:

…I shall not answer it. Indeed, I have already torn it up. Some day, when the summer satrap of the Snooper State comes to ask me why I refuse to contribute my share of statistics to the national numbers game, I shall call for my lawyer. For my house claims protection under the Fourth Amendment: "The right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated, and no Warrants shall issue, but upon probable cause, supported by oath or affirmation, and particularly describing the place to be searched, and the persons or things to be seized." "Go," I shall say, "and report to your Snooperiors! Tell them that I shall resist this unreasonable search! I plead the Fourth!"

Summoned before a Grand Jury in New York City to explain his refusal to fill out the Long Form, Rickenbacker stated, in part: "In the name of interstate commerce, some Government employees think they have the right to pry the roof off my house and watch everything I do there. Let them ask all the questions they please; and let them answer who are disposed to answer, and keep silent who wish to keep silent; that is liberty."

Indeed, it would be a useful exercise to ask where the Constitution grants the federal government authority to compel citizens to provide information about their Hispanic origin, cars, bathrooms, and the rest. And then ask whether any constitutional rationale for the Long Form – interstate commerce, the general welfare, or whatever – could not be used to justify asking ever more outlandish questions about our private lives. If we do not draw the line where the Framers drew it, what’s to prevent the government from trying to collect data concerning our cash transactions, our religious beliefs, our political party affiliation, or our sexual practices?

Third, the Long Form encourages government intervention in the economy, social engineering, and "group think." The statistical aggregates generated by the Long Form foster what Nobel Laureate Friederich Hayek called the "pretense of knowledge" – the conceit that one knows enough to improve upon the outcomes of the marketplace. Armed with the kind of information the Long Form provides, politicians and bureaucrats invariably discover statistical disparities in the incomes, employment levels, educational credentials, and so on among men and women, Hispanic and non-Hispanic, White and Black. Without further ado, they conclude that these disparities reveal "market failures," "glass ceilings," racial discrimination, or, at the very least, "unmet needs." They then enact fiscal and regulatory schemes to remedy the alleged market imperfections and achieve "social justice." Whatever good these programs accomplish, they also typically rob Peter to pay Paul – especially if Paul is a big-time campaign contributor, a mobilized voting bloc, or a powerful Washington lobby.

Statistical disparities among different descriptions of citizens are inevitable in every society. In a free society, they are the result of human action, though not of human design. Furthermore, statistical disparities may conceal more than they reveal. For example, comparing the difference in earnings between the bottom and top quintiles from one decade to the next may create the appearance of a growing income "gap." In reality, this may merely mean that more young people have entered the workforce in the later decade. More importantly, most of the people in the bottom quintile in one decade are in higher quintiles in the next decade. They have not fallen further behind; their economic situation has improved. In addition, many people in today’s bottom quintile enjoy a material standard of living comparable to that of the middle class of 40 or 50 years ago. The statistical snapshot conceals the reality of an upwardly mobile society. The alleged problem of a growing income gap turns out to be a statistical artifact, an illusion.

The mentality that views such disparities as inherently suspect and requiring remedial government action is the root of many entitlement programs, all affirmative action mandates, and numerous regulatory programs. It is the Zeitgeist, the ruling spirit, of the Era of Big Government – an era that is far from over. The Census Long Form will help perpetuate that mentality into the 21st Century.

Much recent social commentary laments the decline of a common American identity – a growing fragmentation along ethnic, racial, and gender lines. Such sweeping generalizations about America are perhaps best taken with several grains of salt. The popularity of both Seinfeld and country music would be hard to explain if America were a nation of discrete and insular minorities. Nonetheless, there is probably something to the widely shared perception that the political balance is shifting in favor of Pluribus and against Unum. What is not always recognized, however, is that social fragmentation flourishes in a politicized economy. When government manipulates the outcomes of the marketplace along "class," gender, and ethnic lines, it fosters social discord. When government allocates billions of dollars based on the kinds of demographic categories enshrined in the Long Form, it promotes "group think" and nurtures a grievance industry with a vested interest in fueling social conflict. The Long Form is part and parcel of an approach to politics that tends to obscure from Americans their common patrimony of liberty.

No doubt someone will observe that the Census Short Form also contains questions about ethnic and racial identity. These, too, should be dropped, in my opinion. America has a colorblind Constitution. We should also have colorblind laws and policies.


III.  Assessment of Arguments for Retaining the Long Form

A common argument made by defenders of the Long Form is that many federal programs explicitly mandate the use of census data, others are required to use census data by the courts, and still others depend on census data for program planning, implementation, and evaluation. To make this point painfully obvious, the Census Bureau has compiled detailed lists of the federal laws that explicitly require agencies to use census data. In other words, we are to infer that abolishing the Long Form would cripple the federal government, that it would paralyze major programs and agencies.

This argument makes three dubious assumptions. One is that lawmakers would abolish the Long Form without changing the statutes mandating the use of Long Form data. Another is that in the interval between abolition and the next census in the year 2010, federal agencies would not develop alternative, non-coercive information sources. A third is that all the programs that currently rely on Long Form data advance the public interest and should be maintained in perpetuity. Mr. Chairman, this is not the place to debate the merits of particular programs. However, I would simply note that the Competitive Enterprise Institute, the Heritage Foundation, the Cato Institute, and other free market think tanks believe that civil society, the American family, and the U.S. economy would benefit tremendously from dramatic cuts in the size, power, and scope of the federal government. Furthermore, if it really is the case federal programs cannot run without Long Form data – that is, without coercively violating the privacy of American citizens – then policy makers ought to review the constitutional propriety of those programs.

Another argument is that abolishing the Long Form would incapacitate state and local governments. In an article in which several authors pretend to evaluate after the fact the consequences of Congress’ abolition of the Long Form, Dowell Meyers laments that while some information on income, jobs, education, transportation, and the like was still available from national surveys, "data at the local level simply vanished." Meyers reports that "Private vendors rushed to fill the void, but they competed on price, rather than quality. After all, without census data, how could anyone judge good data from bad?"

Meyers ignores the fact that data collection, analysis and provision are huge and growing sectors of our economy. Given the importance of accurate data to marketing and investment decisions, it is unreasonable to suppose that firms would compete by selling shoddy products at cutthroat prices. Conceivably this might happen for a year or two if the Long Form suddenly dropped off the face of the earth without warning. But if local governments and private data providers had a decade to prepare and adjust, marketplace competition would improve quality while lowering price, as it has throughout the information services sector.

Meyers also overlooks the potential for private regulation of data quality by independent third parties. I see no a priori reason, for example, why former Census Bureau officials could not start the data marketing equivalent of Underwriters Laboratories. Free market competition allows industry standards to evolve – government did not have to mandate the use digital technology in the recording industry, nor the Windows operating platform in the personal computer industry. Finally, I would note that some state and local governments have budgets that exceed that of the Census Bureau; collectively their budgets vastly exceed that of the Commerce Department. If state and local governments are really so keen to keep tabs on people’s bathrooms and ethnicity, the private sector will figure out how to furnish such data via non-coercive methods.

Another argument is that businesses will lack the demographic and income data required for predicting consumer demand in local markets. In the same fanciful article looking back from the year 2003, Marcia Mogelonsky complains that without block-group-level data on income, supermarkets can no longer calculate consumer demand for many grocery items. "Most food shoppers live within a two-mile radius of their primary food store," she observes. But this fact undermines her thesis. Just eyeballing the neighborhoods within a two-mile radius should give the local proprietor a pretty good idea of income ranges and distribution. Proxy data such as housing values – available from real estate brokers or property tax assessors – might be used. Supermarkets could pool their resources to buy marketing data on a discounted basis. If business really needs the kinds of data supplied by the Long Form, then business should pay for it, not the taxpayer. Nor should taxpayers be coerced into providing it, as they are today.

Writing in the same article, Keith Wardell argues that private purveyors of marketing data will do just fine in a world without the Long Form. "Customer lists are fairly easy to enhance without the census. For a mailer’s house file, enhancement with third-party databases, such as INFOBASE, or direct surveys of customers can begin to replace information provided by the Census Bureau," he says. Wardell’s fanciful retrospective is worth quoting at length:

Prospecting for new customers may also be increasingly effective without a census. Marketing information built huge and detailed consumer databases for prospecting in the 1990s. Questionnaire programs grew significantly. These come in many variations, including warranty cards, solo mail, and ride-alongs from companies like Polk/NDL, Shareforce, and Buyer’s Choice. In addition, a number of marketing information companies began combining customer files from many different mail-order companies in order to better understand consumers. These include Abacus, DMI’s SmartBase, and Direct Tech’s Z24….As these databases grew, they provided more accurate targeting information than was provided by the census long form. The use of the data provided by the census was already on the decline by the time the 2000 census became available.

Mr. Chairman, I do not pretend to be an expert in the evolutionary dynamics of the data marketing industry. But common sense suggests that if Long Form data are so valuable to business, then abolishing the Long Form will confer a boon on data marketing companies. Demand for their products and services will soar. Cutting government will create private sector jobs.


IV. Conclusion

That consideration, however, is of secondary importance. To recapitulate, abolishing the Long Form will allow the Census Bureau to concentrate on its constitutional mission to produce an accurate enumeration of citizens for purposes of apportioning representatives in Congress. It will reduce government’s coercive intrusion into the private lives of American citizens. And it will de-politicize or privatize the collection of data, perhaps deflating somewhat the pretensions of the grievance mongers and would-be central planners. A national debate over abolition of the Long Form would also have immense educational value, reminding Americans how far our government system has strayed from its original constitutional design.

1August Gribbon, "Compilers of 2000 census face diminishing returns," The Washington Times, May 18, 1998, A12. 2Id. 3James Bovard, Lost Rights: The Destruction of American Liberty (New York: St. Martin’s Press, 1994). 4William F. Rickenbacker, "The Fourth House," National Review, May 21, 1960, p. 325. 5"How Many Bathrooms Have You?" National Review, December 31, 1960, p. 399. 6Rickenbacker, "Fourth House," p. 326. 7John T. Wenders, "Statistics: A Vehicle for Collectivist Mischief," The Freeman: Ideas on Liberty, June 1998, p. 326. 8Wenders, Id. 9See Bureau of the Census, Planning for the Census 2000: Questions Planned for Census 2000, Federal Legislative and Program Uses, March 1998. 10See, for example, CEI, The Dirty Dozen: Soft Targets for Elimination at Energy, Interior, & EPA, December 1994, and The Baker’s Dozen: More Soft Targets for Elimination, April 1995; Scott Hodge, ed., Rolling Back Government: A Budget Plan to Rebuild America (The Heritage Foundation, 1995); Cato Handbook for Congress: 105th Congress, 1997, especially "Federal Agencies: The Abolition Agenda," pp. 145-179. 11"Life Without the Census by census data users," American Demographics, October 1995. 12On this general subject, see Yesim Milmaz, "Private Regulation: A Real Alternative for Regulatory Reform (Cato Institute Policy Analysis, No. 303), April 20, 1998. 13Id. 14Id.