After Donald Trump’s presidential victory, many left-leaning politicians and commentators started considering what sort of message would help them regain political power. Given the clear racial subtext to the 2016 campaign—Trump stereotyped Mexican immigrants and Muslim families, and Hillary Clinton spoke openly about systemic racism—and who ultimately won, some concluded that Democrats should stop using so-called “identity politics” to try to win elections. The argument went that this approach would alienate some largely white swaths of the country that would be crucial to the party’s future electoral success.

If politicians heed this call, one potential result could be that they would move toward policies that direct attention away from race and instead focus broadly on economic mobility. The past 100 years, however, hold a few lessons about the consequences of enacting intentionally or unintentionally race-blind policies.

In his book When Affirmative Action Was White, the Columbia University political scientist Ira Katznelson discusses some examples of economics-oriented policy initiatives that favored white people, including many that were ostensibly race neutral. Most of these colorblind policies were developed as part of the Roosevelt administration’s New Deal programs, and were designed to help lift America out of the Depression. Social Security, minimum-wage laws, and the G.I. Bill are all examples of legislation written with language that wasn’t explicitly discriminatory. And yet, these policies did not work for everyone: It was primarily white Americans who were able to benefit from them once they were put in place.

Take the G.I. Bill. Katznelson notes that it promised veterans returning home from World War II a number of benefits: access to low-interest home loans, money for education, and job training. The letter of the law did not favor veterans of one race or another, but in practice, several things prevented black Americans in particular from getting their share of benefits. First of all, the bill was administered by states, not the federal government. This meant that Southern states, which had kept up a system of legal segregation, could require that black veterans seeking college degrees enroll at historically black colleges, which were almost always underfunded and had few of the resources that primarily white colleges did. In fact, Katznelson notes, some Southern congressmen supported the bill precisely because of the presence of such loopholes.

Likewise, the job-training component of the G.I. Bill was hampered by the fact that it was implemented on the state level. The program’s administrators, most of them white, tended to steer black veterans away from higher-paying jobs in carpentry, woodworking, photography, or electrical work and instead toward training programs for lower-paying service jobs, such as those in dry cleaning or tailoring. This happened despite the fact that black newspapers at the time widely reported that black veterans were just as eager as white ones to take advantage of the benefits promised to them by the government.

The G.I. Bill also provided for veterans to get affordable home loans with low down payments, but these too ran up against the realities of discrimination: Banks, which were responsible for actually making the loans, were (and indeed, still are) usually much more inclined to grant them to white people than black people. Many white vets availed themselves of the G.I. Bill-sponsored loans, which facilitated a wave of mostly white homeownership and the growth of the mid-20th-century middle class. Additionally, as segregation was still commonplace, even when black veterans did secure loans, they were often only for houses in predominantly black areas where property values were lower.

In all of these cases, the fact that the law wasn’t written with preexisting racial imbalances in mind meant that many non-white veterans were left without access to classic American paths to upward mobility. And, on top of that, the law didn’t account for another danger of colorblind policies, which is that failing to provide for existing inequalities leaves room, down the line, for people to enforce the law unevenly. The end result of all this is what Katznelson describes as affirmative action for whites. This was not because Congress wrote legislation explicitly intended to disenfranchise veterans of color; rather, the G.I. Bill—like many of the other policies Katznelson describes in his book—was written as race-neutral and specifically stated that all veterans were eligible. As Katznelson shows, the law didn’t fully deliver on its promise because it didn’t devote any special attention to the racial dynamics that undergird employment, homeownership, and education.

In another study, Leland Saito, a sociology professor at the University of Southern California, offers more-recent examples of how ostensibly race-neutral policies can end up marginalizing people of color. While When Affirmative Action Was White focuses on federal policy, Saito’s 2009 book The Politics of Exclusion has its eye on the local level, considering the results of cities’ and towns’ economic-development programs. These programs, Saito argues, are a good example of race-blind policies designed to bring about widespread improvement. However, he shows that an inattention to race precludes local officials from seeing how such policies can overlook nonwhite groups that aren’t well-represented in politics.

As one example, he brings up efforts in the late 1980s to revitalize an area of downtown San Diego called the Gaslamp Quarter, one of the city’s major entertainment destinations. As city planners launched a redevelopment plan and surveyed the area—which was home to, among other cultural landmarks, the Chinese Mission, a social center with lots of significance to Chinese workers who had lived in the city over the years—they found little they considered worth preserving. Concerned Chinese Americans felt compelled to form the Chinese Historical Society, which ultimately persuaded the developers to save the Chinese Mission. Saito’s focus on Chinese Americans in San Diego highlights the ways Asian Americans, often stereotyped as a model minority, find that their particular racial experiences are frequently overlooked and ignored. And more generally, Saito makes the point that if a sensitivity to race and ethnicity is not baked into policy, there is a real possibility that what’s important to nonwhite groups may be (quite literally) bulldozed.

Saito also shows how the same revitalization initiative affected many African Americans in San Diego who treasured the Douglas Hotel, a major entertainment venue and the lone hotel in the city’s center that would accommodate blacks when segregation was explicitly legal. As with the Chinese Mission, politicians and legislators using a colorblind, big-picture economic approach did not see how important certain historic landmarks were to communities of color and shaped the city’s cultural fabric. In this case, the Douglas Hotel was razed in 1985 to build residential and commercial properties.

Economic development policies are of course different from the G.I. Bill, in the sense that they are designed to improve cities overall rather than to offer opportunities to specific groups of people. But it isn’t hard to imagine how policymakers who consider abandoning identity politics might pursue these sorts of initiatives in urban areas, unaware of the impact these efforts could have on communities of color. Even policies that appear to have nothing to do with race can still lead to outcomes that disproportionately harm entire racial groups.

Perhaps the most pernicious examples of race-blind policies that have had racially disproportionate outcomes can be found in the criminal-justice system. In a 2011 article in the journal Crime and Delinquency, Traci Schlesinger, a professor of sociology at DePaul University, argued that at the state level, mandatory minimums and sentencing enhancements disproportionately affect black men. Schlesinger notes that while these policies were designed to standardize punishments and minimize any disparities that might come from juries in various locations assigning wildly different punishments for criminal activity, in practice they still lead to arrests and charges that are more likely to target blacks.

Both mandatory minimums and sentencing enhancements require harsh punishments for crimes committed in or near public spaces, with firearms, or by people with prior convictions. But Schlesinger notes that due to long-term, ongoing patterns of racial segregation and economic disadvantage, blacks are more likely to live in or near public spaces like public housing. And despite exhibiting a rate of drug use that is roughly comparable to that of their white counterparts, blacks are more likely to be arrested for drug-related criminal activity, making them more likely to have prior convictions that require mandatory minimum sentences. These facts lead Schlesinger to conclude that “racial disparities are primarily produced and maintained by colorblind policies and practices.” Essentially, because these sentencing policies didn’t incorporate disparities that already existed, they inadvertently produced more of them.

Overall, the work of Saito, Katznelson, and Schlesinger offers a cautionary note about what can happen when those in charge of making policy abandon identity politics and ignore entrenched inequalities based on race, gender, ethnicity, and other categories. Regardless of whether a more colorblind platform would broaden the Democratic base in the ways the party needs—and it’s not clear if that would even happen—there are copious examples of how that strategy would stifle progress on reducing these stubborn disparities.