Politics

When the Justice Department announced in early August that it intended to investigate the admissions practices of selective universities, the reaction was mixed. Many supporters of the Trump Administration celebrated the news as a signal that white applicants might no longer be rejected in favor of those from historically underrepresented groups. Meanwhile, those who have long defended affirmative action policies saw the move as yet another indication that the Administration was working to turn back the clock on civil rights, and that affirmative action policies in general were doomed.

The simple truth of the matter is that affirmative action programs have been highly controversial from the time they came into wide use in the 1970s. An overview of affirmative action from 1978 forward shows that neither the law nor the arguments for and against affirmative action has changed very much at all. What has changed over the last 40 years are the tactics used by activists, administrators and policymakers on both sides of the issue.

In 1978, the Supreme Court announced a confusing decision in Regents of the University of California v. Bakke, 438 U.S. 265, which upheld the principle of affirmative action in university admissions while striking down the quota-based system in use at the University of California-Davis medical school. Bakke was a white Vietnam war veteran with an outstanding service record and impressive MCAT scores who tried for several years without success to get into medical school. He was rejected everywhere, mainly on account of his age — Bakke was in his early 30s, which was considered in those days to be too old for medical school.

Bakke finally applied to UC-Davis, which at the time reserved 8 of its 100 admission slots for minority applicants. He was rejected there, too. He sued the school on the grounds that its quota system violated his right to equal protection under the law under the 14th Amendment. The California Supreme Court eventually agreed with him, and he was admitted to the Davis medical school.

The University of California Regents then sought further review in the U.S. Supreme Court, which held that the practice of reserving designated class space for any particular group was a violation of Bakke’s equal protection rights, but that more nuanced affirmative action programs that reasonably seek to redress historical patterns of discrimination were permissible. This opened the door toward widespread adoption and expansion of affirmative action policies that considered race, sex or other characteristics as just one factor among many in university admissions decisions, even throughout the Reagan and Bush administrations in 1981-93.

Affirmative Action Pushback and Workarounds

The first successes in the fight to dismantle affirmative action programs came not through legislation or the courts, but through the ballot box. Voter initiatives and ballot propositions to reject race or ethnicity as legitimate factors in university admissions began with California’s Proposition 209, which was approved by voters in 1996. Similar measures were soon passed afterward in Michigan and Washington state. Nebraska, Arizona and Oklahoma have followed suit since then. Texas eliminated race as an admissions consideration by court order, Florida abolished racial considerations in affirmative action by executive decree, and in 2011, the New Hampshire legislature passed a bill to the same effect.

Most of these states have other programs in place to achieve goals similar to those of affirmative action programs. For example, California, Texas and Florida have so-called percent plans, which offer automatic admission to a state college or university if a student graduates in the top 4 percent (California), 10 percent (Texas) or 20 percent (Florida) of the student’s graduating high school class.

Other conditions and restrictions concerning academic rigor, course distribution and similar factors apply under each of these programs. In California and Florida, the offer applies to state institutions in general, not necessarily the student’s first choice. In Texas, however, the student actually gets to choose the school.

At UT-Austin, 75 percent of each freshman class is reserved for students who qualify under the Texas 10-percent plan, which makes competition for the remaining places extremely tight. In fact, the 75-percent cap means that the 10-percent qualifying rule might only cover the top 7 to 9 percent of graduating students in a given year.

The percent plans widen the admission opportunities for motivated students from disadvantaged backgrounds and who attend minority-dominated high schools to gain admission to colleges and universities — instead of competing with a broader applicant pool largely composed of students with much better access to academic resources, they’re competing with their own classmates, who frequently have similar ethnic, family income, and academic support characteristics.

Another strategy, such as the one in use at the University of Washington, is to invite consideration of socioeconomic factors and hardship narratives not explicitly linked to race in order to achieve substantially the same goals as a race-based affirmative action program.

Some schools, such as the University of Georgia and Texas A&M University, have abandoned longstanding legacy preference practices, under which children or grandchildren of alumni are given special consideration in admissions decisions. This practice obviously tends to undermine the goals of affirmative action by reinforcing established elite networks.

Are Asian-American Applicants Helped or Harmed by Affirmative Action?

One relatively new element in the affirmative action debate concerns Asian- American students. In 2015, a coalition of 64 Asian advocacy groups asked the Justice Department to investigate whether the nation’s most selective universities were discriminating against Asian-American applicants. This is the complaint that the Department recently revived.

While Asian-Americans make up less than 6 percent of the population nationally, they represent 22 percent of Harvard’s incoming freshman class. How can that be discrimination? The advocacy groups argue that the most selective universities apply an arbitrary cap to Asian-American admissions, and thereby reject Asian candidates who would have been admitted if they were of any other ethnicity. One 2009 study found that Asian-American applicants need to score 140 points higher on the SAT than other groups in order to be considered on an equal footing.

Careful and limited consideration of race continues to be permissible in university admission decisions, at least where state ballot measures or legislation haven’t prohibited affirmative action. This principle was upheld last year by the Supreme Court in Fisher v. Texas. Decided by a 4-3 vote after the death of Antonin Scalia, the Fisher decision was argued and determined on grounds not much different from those raised in the Bakke case almost 40 years earlier.

All of which means that our nation continues to debate the fairness of potentially discriminating against one group in order to prevent discrimination against another group. While the equal protection arguments raised by opponents of affirmative action make a certain intuitive sense, Justice Blackmun’s answer in his Bakke opinion (438 U.S. at 407) continues to resonate:

I suspect that it would be impossible to arrange an affirmative-action program in a racially neutral way and have it successful. To ask that this be so is to demand the impossible. In order to get beyond racism, we must first take account of race. There is no other way. And in order to treat some persons equally, we must treat them differently. We cannot – we dare not – let the Equal Protection Clause perpetuate racial supremacy.

Required every ten years as prescribed in the Constitution, the next U.S. Census is scheduled to take place on April 1, 2020. Data generated by the census dictates each state’s representation in Congress for the next ten years. Each state’s electoral votes can increase or decrease accordingly. The redistricting of congressional and legislative districts within states also depends on census data.

Federal spending in the range of $600 billion is proportionately distributed among the states based on census information. Private industry uses census data for a variety of planning, marketing and growth-management decisions. Meanwhile, as the U.S. population continues to grow, the task of counting 325 million-plus Americans is becoming more complex and more expensive.

How will the census fare under the new Trump administration? So far, the outlook is not too bright. While the White House has taken no steps to interfere with the work of the Census Bureau, it has done little or nothing to support it, either. The top two leadership positions in the Bureau are now vacant, and the president has yet to nominate anyone to head the census. As of August 2017, the 2020 Census faces significant problems in the following areas:

Insufficient Congressional Funding

The average cost of counting Americans has risen in constant dollars from $16 per household in 1970 to $92 in 2010, mainly because Americans in general increasingly have been ignoring the census questionnaires that are sent to every household. The mail-back response rate fell from 78 percent in 1970 to 63 percent in 2010, according to Government Accountability Office (GAO) figures. Anyone who does not complete and return their census form will be visited by a census enumerator, up to six times if necessary. And that’s expensive.

On the face of it, the Census Bureau fared relatively well in the proposed White House 2018 budget, because it was granted a slight increase of $50 million over its 2017 funding — a sharp contrast to the deep cuts imposed on most other domestic government agencies and functions. Previous budget projections would have provided an additional $200 million to the Census Bureau in 2018. The 2020 Census will almost certainly be forced to count a larger and less-responsive population with fewer resources.

According to the GAO, the 2010 Census was the most expensive American census in history. It cost about $12.3 billion, which reflected a constant-dollar increase of 31 percent over the cost of the 2000 Census. Right now, the Census Bureau is committed to operating under a congressional mandate that the 2020 Census will cost no more than the 2010 count, or about $13 billion in constant dollars. The Bureau hopes that it can find about $5 billion in savings through technology-driven efficiencies, but it has shown little actual ability to do so. Developed in October 2015, the Bureau’s cost estimate for the 2020 Census will be updated this summer, and it seems likely that the cost will ratchet upward.

Funding problems have already begun to affect the Census Bureau’s work. The 2018 End-to-End Test, a comprehensive dry run of the 2020 Census, was supposed to begin next summer in Providence County, R.I., Pierce County, Washington, and a nine-county rural area in southeastern West Virginia. This plan was recently scaled back to full execution only in Rhode Island, due primarily to the $150 million shortfall between the Obama and Trump administrations’ projected census funding for fiscal 2018. Field testing of new methodologies scheduled for 2017 was cancelled in October 2016.

Unstable Leadership

The Director of the Census Bureau is appointed by the president and confirmed by the Senate. Just a few days after his testimony before a House subcommittee in May 2017, Census Director John Thompson resigned his position. The deputy director, Nancy Potok, left her Bureau office in January to accept another federal appointment. Commerce Secretary Wilbur Ross named Jon Larmin to serve as acting director and Enrique Lamas as acting deputy director on June 29. Both interim appointees are long-term Census officials who have served in various capacities with the Bureau, but neither has much job security. Either the President or the Secretary of Commerce could replace either or both officials on little notice.

If the President or Secretary names new officials to replace the current Census leadership with little time remaining before the 2020 Census, the Bureau’s current challenges could be compounded by a loss of continuity, decline in morale, and sudden shifts of emphasis — all of which could affect the reliability and cost of the 2020 Census results.

Concerns Over Technology

Congress apparently expects the use of high-tech data gathering methods to help cut the costs of performing the 2020 Census. Internet census returns have been considered as a way to reduce the expense of field operations, but it’s hard to imagine that declining public confidence in online data collection will result in higher or less-expensive return rates from American households. Hacking scandals in banking, retail, entertainment, politics and other areas are becoming more frequent, not less, and public skepticism over cybersecurity will represent an important barrier to full participation.

It is not yet clear that the 2020 Census is committed to substantial public participation online, but it seems doubtful that an option to return the census questionnaire over the internet will significantly increase the return rate over 2010’s 63 percent. Meanwhile, the Census Bureau’s current cost overruns primarily involve the difficulty in developing electronic data collection methods that are accurate, effective, scalable and secure. To the greater extent that Americans fail to respond to census questionnaires either by mail or online, the more dependent the Census will be on the highly expensive task of having enumerators go from door to door in order to count everybody.

The GAO has identified the 2020 Census as a government program at high risk of vulnerability to fraud, waste, abuse, or mismanagement, or most in need of transformation. We’ll find out soon whether this assessment has generated enough concern to result in the corrective actions necessary to make our next census a success.

When President Trump issued his May 11 executive order creating the Presidential Advisory Commission on Election Integrity, the move was hailed by some as an important step toward reducing the risk of voter fraud in American elections. Other commentators, pointing to the absence of documented voter fraud beyond a few isolated cases, considered the new Commission to represent little more than a tool for promoting stricter voter registration requirements that would disproportionately affect younger voters, low-income voters, naturalized voters, and voters of color.

Barely two months after its creation, the Commission finds itself embroiled in controversy on multiple fronts. Practically none of the states requested to provide comprehensive voter information to the Commission have done so by the original July 14 deadline, and at least six legal or administrative complaints have been filed to challenge the Commission’s creation and functions.

On July 10, the Commission administrator asked states not to provide the requested data until a United States District Court rules on a motion for temporary relief based on the Commission’s alleged violations of the E-Government Act of 2002, which requires federal agencies to conduct a Privacy Impact Assessment for any data systems that contain personally identifiable information. The motion was filed by the Electronic Privacy Information Center (EPIC) on July 3.

In the previous week, Commission Vice-Chair Kris Kobach requested that the secretaries of state for each state and the District of Columbia provide the following information with respect to each registered voter in each jurisdiction:

• Full name
• Address
• Dates of birth
• Political party
• Last four digits of Social Security numbers
• A list of the elections voted in since 2006
• Felony convictions
• Voter registration information from other states
• Military status
• Whether voters lived overseas

EPIC’s motion for temporary relief argued that release of this voter information would “increase the risks to the privacy of millions of registered voters—including in particular military families whose home addresses would be revealed—and would undermine the integrity of the federal election system. Further, the request for partial Social Security Numbers that are often used as default passwords for commercial services, coupled with the Commission’s plan to make such information ‘publicly available,’ is both without precedent and crazy.”

The requested individual voter data was to have been sent by the states to the U.S. Army’s SAFE website, which in turn was to have forwarded the data to White House servers. The Army itself is a defendant in a recent lawsuit filed by Public Citizen, Inc., which alleges that using the Army website for the Commission’s work violates the Administrative Practices Act and the Privacy Act of 1974.

It’s unclear why the Commission started its work with this blanket request for individualized voter information data. The request appears to be outside the scope of the stated purpose of the Commission to (1) identify laws and practices that strengthen public confidence in federal voting procedures, (2) identify laws and practices that undermine confidence in such procedures, and (3) identify the vulnerabilities that could lead to improper or fraudulent registration and voting in federal elections. Once the Commission delivers its report to the President on these three points, its work will be complete and it will dissolve.

In the near term, however, the Commission’s activities are likely to be bogged down in litigation and noncooperation from state election officials. It remains to be seen whether the Commission on Election Integrity will backtrack and demonstrate its compliance with the Privacy Act, the E-Government Act, the Federal Advisory Commissions Act, and the Administrative Procedures Act before renewing or refining its request for sensitive individual voter data.