Thursday, the Voting Rights Act of 1965 (VRA) turns 50. This landmark act was designed to ensure that states could no longer prevent minority citizens — African Americans, in particular — from voting. Journalists and academics are noting both how far we’ve come and how many challenges remain to ensuring that minority voters can cast their ballots.

Exactly how has minority participation in elections changed since the Civil Rights era?

African Americans can (and do) vote at far higher rates than they did in the decade before the Voting Rights Act, according to the evidence. But Latino and Asian-American turnout remain so low that our elections may be less representative of the racial/ethnic mix of the country than they were 50 years ago.

Here are three trends you might not know about minority voting over the last half century.

1. Black turnout and white turnout are about equal in presidential elections today. That’s a dramatic shift from the 1950s and 1960s.

Until the early 1960s, authorities in the South used various methods to keep black people from taking part in politics. So it’s no surprise that in the 1952 and 1956 presidential elections, turnout among voting-age African Americans was below 20 percent in southern states. Grass-roots activism by African-American groups brought those numbers up substantially by the early 1960s — but the VRA solidified these gains and brought federal oversight that forced change from the remaining Southern states and counties that refused to allow African Americans to register and vote.

After the VRA was implemented, the South’s racial voting patterns began to look much more like those in the rest of the country. The states and counties that had been most intransigent in stopping blacks from voting were specifically required to get all voting processes approved in advance by the U.S. Department of Justice, a feature called “preclearance,” which the Supreme Court invalidated in Shelby County v. Holder in 2013. By the mid-1980s, between 55 and 65 percent of both black and white citizens voted, in the South and non-South. In the South, that meant a 5 to 10 point gap between the two groups – which was far less than the 20 to 40 point gap found before the VRA.

Now jump to recent times. In the last two presidential elections, in 2008 and 2012, black turnout and white turnout were virtually equal — even in the South. Racial differences in overreporting may explain some of the recent increase in black turnout relative to whites. Barack Obama’s historic candidacy and effort to mobilize black voters probably plays a role as well.

In short, we have witnessed a monumental change in black voting over the past 50 years.

2. Black turnout continues to lag white turnout in midterms, and Latino and Asian turnout is even lower.

The story shifts when we include turnout in midterm elections. In 2010 and 2014, black turnout still lagged about 5 percentage points behind whites. The gap was even larger in 2006. While we certainly see signs of a closure of the turnout gap similar to that found in presidential elections, it is misleading to suggest that “the racial gap in electoral participation has by now disappeared.”

In 1975, the VRA was amended to offer protections to Latino and Asian American voters, especially those whose primary language was other than English. However, Latino and Asian-American voting rates have remained 15-20 percentage points behind non-Hispanic whites, even in presidential elections, since those groups’ voting started to be counted. (The federal government’s Current Population Survey, widely considered to have the best available data, did not identify Latinos before 1972, and does not have information on Asian Americans as a group prior to 1990).

In fact, Latino and Asian turnout may actually be getting worse over time, both compared to percentages of white citizens voting and (at least in the midterms) in raw percentages.

3. Because such low percentages of Latinos and Asians vote, today’s voters represent our nation’s diversity less accurately than in the 1950s and 1960s.

In the 1950s, the nation was approximately 90 percent white and 9 percent black, with very small percentages of other racial minorities. (As noted above, Latinos were generally not counted separately until the 1970s). Because black citizens were kept from the polls, the electorate looked substantially “whiter” than it should have. For example, in 1952 the voting population was 95 percent white and 4 percent black, meaning the electorate was five percentage points whiter than the adult population.

Voters started looking more like the population as a whole in the years leading up to, and immediately following, passage of the VRA. In fact, the 1964 and 1968 presidential elections saw a voting population that closely mirrored the racial composition of the nation. As noted above, in 2008 and 2012, black voters were slightly overrepresented compared to the nation’s population.

But as we know, the United States has grown more ethnically diverse, with far more Latinos and Asian Americans than in the past. In 1988, the voting eligible population was roughly 84 percent white, 11 percent black, and 5 percent Latino, while voters were 86 percent white, 10 percent black, and slightly less than 4 percent Latino. Latino voters were underrepresented, but since there were so few Latino citizens anyway that this might not have had a substantial effect on national or most state-level elections.

The gap is much bigger now. In 2014, for instance, citizens eligible to vote were 72 percent white, 12 percent black, 12 percent Latino, and 4 percent Asian. But those who actually voted were 78 percent white, 12 percent black, 7 percent Latino, and 3 percent Asian.

As a result, white voters were even more overrepresented in 2014 than they were in 1952–despite the dramatic changes in protections for minority voting rights. Though the non-Hispanic white population makes up an ever-smaller share of the population, low Latino and Asian turnout may mean that whites continue to have outsized influence in elections despite the demographic shift.

Fifty years on, much has changed regarding minority participation in the electoral process. During the 1960s, attempts to block African Americans from voting were ludicrously easy to find. Today those barriers are difficult to identify. And yet Latino, Asian American, and (more often than not) black turnout continues to lag that of whites. We should continue to assess the causes and consequences of an electoral process whose participants are substantially less diverse than the population as a whole.

Bernard L. Fraga is an assistant professor of political science at Indiana University. He studies electoral politics and institutions, with a focus on the role of race and ethnicity in American elections.