Do 40‐​Year‐​Old Facts Still Matter?: Long‐​Run Effects of Federal Oversight under the Voting Rights Act

January 16, 2019 • Research Briefs in Economic Policy No. 146
By Desmond Ang

The Voting Rights Act (VRA) of 1965 has been hailed as one of the “greatest legislative achievements of the Civil Rights Movement.” Passed months after the Selma to Montgomery marches, the Act prohibited the denial or abridgement of “the right to vote on account of race or color.” The effects of the VRA on minority enfranchisement were immediate. Between the 1964 and 1968 presidential elections, black voter registration rates increased 67 percent among Southern states.

The Act achieved this through two principal mechanisms. The first was the prohibition of literacy tests, which were used throughout the Jim Crow era to disenfranchise Southern blacks. Section 5 of the VRA established a second and more controversial mechanism: a federal oversight process commonly referred to as “preclearance.” Jurisdictions subject to preclearance (henceforth called “covered jurisdictions”) were prohibited from implementing any new electoral rule without first obtaining federal approval. While preclearance’s geographic purview was limited only to areas that met certain historical criteria, the scope of its protections was expansive and encompassed all future changes affecting voting in those areas. Thus, preclearance restrictions, which have been called “the most effective means of preventing racial bias in voting,” were designed as a broad prophylaxis against voter discrimination, shifting onto covered jurisdictions the burden of proving ex ante that new voting rules did not have a “discriminatory purpose” and would not have a “discriminatory effect.”

Since its inception, preclearance oversight has been alternately praised and criticized as “extraordinary legislation otherwise unfamiliar to our federal system.” These arguments came to a head in 2013 in Shelby County v. Holder, in which the Supreme Court ruled that continued coverage based on historical—rather than current—measures of discrimination is unconstitutional. As a result, until and unless Congress enacts a new coverage formula, previously covered jurisdictions are no longer subject to federal oversight.

Immediately following the Shelby ruling, lawmakers in several previously covered areas enacted controversial new voting changes, many of which have been challenged in federal courts. Alabama, Mississippi, North Carolina, and Texas introduced restrictive voter ID requirements, while Florida, Georgia, and Virginia sought to purge their voter rolls of thousands of eligible minorities. Though Republicans have justified these measures as necessary to combat widespread voter fraud, Senate Democrat Chuck Schumer denounced them as a “clear front for constricting the access to vote [for] poor Americans … African‐​Americans and Latinos.” Underpinning this partisan divide is the common belief that the minorities most impacted by restrictive voting rules lean heavily Democratic. Indeed, President Donald Trump claimed that, of the “millions” of allegedly illegal ballots cast in 2016, “none of ’em come to me. They would all be for the other side.” Given America’s growing minority electorate, the legal fate of these voting laws could have lasting implications for future elections.

Despite their relevance to ongoing policy debates, the specific effects of preclearance have never been estimated. While researchers have examined the VRA’s impact on turnout and representation, these studies focus on the 1965 implementation of the Act and are thus unable to disentangle the effects of preclearance from the simultaneous abolition of literacy tests, which were among the most discriminatory tools ever employed in the U.S. election system and are unlikely to ever be reinstated. Furthermore, those studies as well as the broader literature on enfranchisement focus on policies designed to alleviate specific, existing barriers to voting—such as the elimination of literacy tests or the expansion of suffrage rights.

Preclearance restrictions differ fundamentally from these interventions. Rather than targeting individual voting barriers already in use, federal oversight was designed to restrict the implementation of any and all new discriminatory measures. Understanding the implications of these blanket protections is especially relevant considering the strategic manipulations that local election officials engage in to maintain power. Broad preventative oversight encompassing the universe of potential voting changes may be the most effective means of curbing discrimination in the United States, where electoral rule‐​making is highly decentralized and opaque.

My research seeks to better understand the effects of such oversight. I examine the geographic expansion of coverage under the 1975 revision of the VRA to estimate the causal impact of preclearance on county‐​level voter turnout and Democratic vote share from 1960 to 2016. Unlike the 1965 VRA, which was reverse engineered by Congress to target Southern states that employed literacy tests, the 1975 coverage formula relied on noisy measures of voter turnout and minority population share to determine which counties were subject to preclearance. Thus, application of the 1975 formula resulted in heterogeneity of coverage within states throughout the country, subjecting 283 counties across nine states to federal oversight. I am able to exploit this heterogeneity to precisely estimate the policy’s effects and to demonstrate its plausible exogeneity.

I find that preclearance restrictions led to gradual and significant increases in voter participation and that these gains persisted for over 40 years, bolstering turnout by 4 to 8 percentage points in recent elections. Examining state‐​level turnout by race, I demonstrate that these effects were due entirely to increased participation among minorities, who were 30 percent more likely to vote in the 2012 election as a result of federal oversight. Analyzing electoral rules data, I show that municipalities subject to voter protections were significantly less likely to employ “winner‐​take‐​all” election systems, which are commonly believed to dilute minority voting power. Combined with heterogeneity analysis demonstrating larger effects among areas with greater historical discrimination, these results suggest that gains in turnout were the result of reduced voter discrimination as opposed to other demographic or political factors.

Surprisingly, I find that preclearance coverage led to significant and immediate decreases in the share of Democratic votes cast. These estimates are large—averaging 3.2 percentage points across post‐​treatment elections—and exceed the 1992 and 1996 presidential margins of victory in the covered states of Texas and Arizona. Using historical survey data, I show that this rightward shift was driven by increased Republican support among whites. In particular, whites who were opposed to the Civil Rights movement were 30 percent more likely to identify as Republican following the implementation of preclearance in their counties. In further demonstration of the political controversy surrounding these measures, I find that newspaper mentions of the VRA spiked sharply in covered areas beginning in 1975, particularly among those papers that had endorsed President Richard Nixon, whose Republican administration had sought to abolish preclearance. Taken together, these results provide strong evidence that the enactment of minority voter protections triggered a massive political realignment among the white electorate.

Finally, to investigate the ramifications of the Supreme Court’s 2013 decision to overturn preclearance, I examine returns from the 2016 presidential election. I find that areas that were newly freed from federal oversight experienced a significant decrease in turnout of 1.5 percentage points, the single largest year‐​to‐​year drop in the 40‐​year sample period. While white turnout in these areas remained unchanged, minority participation dropped by nearly 4 percent. These results bolster claims regarding the discriminatory nature of recent election law changes and provide early evidence that the Shelby ruling may jeopardize decades of voting rights progress.

In the majority opinion, Chief Justice John Roberts struck down preclearance on the grounds that “a [coverage] formula based on 40‐​year‐​old facts” has “no logical relation to the present day.” In contrast, I find evidence not only that preclearance coverage led to increases in minority participation that persisted for more than four decades, but also that these gains were quickly eroded after its rescission. To the extent that the future of the VRA hinges on the formulation of new coverage criteria relevant to the “present day,” understanding these effects and the role they played in shaping the current political landscape is critical to Congress’ ability to craft meaningful legislation capable of ensuring the voting rights of American citizens.

This research brief is based on Desmond Ang, “Do 40‐​YearOld Facts Still Matter? Long‐​Run Effects of Federal Oversight under the Voting Rights Act,” American Economic Journal: Applied Economics (forthcoming), https://​www​.aeaweb​.org/ articles?id=10.1257/app.20170572&&from=f.

Media Name: rb146-cover.jpg

Download the Research Brief in Economic Policy

About the Author

Desmond Ang, Harvard University Kennedy School of Government.