The recent unrest in Baltimore, Ferguson, and other cities is puzzling in one important respect. Unlike in earlier eras, when African Americans’ political exclusion drove them to protest, blacks today are as likely to vote as whites and are well represented at all levels of government. The mayor of Baltimore and a majority of its city council are black. So are forty-five members of Congress—an all-time high. And, at the other end of Pennsylvania Avenue, so is the current occupant of the White House. Why all the turmoil, then, at a time when blacks—finally—seem to be enjoying the fruits of American democracy?

One answer is that the appearance of black political clout is deceiving. Despite their gains in participation and representation, blacks continue to fare worse than whites in converting their policy preferences into law. This poor performance is more revealing than statistics on turnout or black electoral success. And even though its causes remain mysterious, it is very much a rationale for frustration with the status quo.

In a recent study, I analyzed group political power at the federal and state levels. At the federal level, I relied on a remarkable database compiled by Princeton political scientist Martin Gilens. It includes responses to thousands of survey questions from the last few decades. Crucially, it also tracks whether each policy referred to by a question was adopted by the federal government over the next four years. At the state level, I measured people’s ideologies using exit polls that asked whether they are liberal, moderate, or conservative. And I assessed state laws using an index of overall policy liberalism created by another pair of scholars.

At both levels, I found that blacks hold much less sway than whites. For example, a federal policy with no white support has only a 10 percent chance of being enacted, while one with universal white support has a 60 percent shot of adoption. But while a proposal with no black support has a 40 percent chance of becoming law, one enjoying unanimous approval has only a 30 percent probability of enactment. In other words, as support for a policy rises within the black community, the odds of it being achieved actually decline.

Likewise, whether most black voters are conservative or liberal, state legislative outcomes barely budge. But vary the views of white voters to an equivalent degree, and a state’s policies go from looking like Alabama’s to resembling Michigan’s, even controlling for black and white population size.

The story is similar for several other groups. The more that women, the poor, or Hispanics support a federal policy, the less likely the policy is to be enacted. Strikingly, as women move from universal opposition to a proposal to universal support, its odds of adoption plummet from 75 percent to 10 percent. Changes in the ideology of female or poor voters also have no effect on state legislative outcomes (although shifts in the views of Hispanic voters do). In contrast, both federal and state laws are acutely sensitive to the preferences of whites, men, and the rich.

These results help explain the rage that has recently coursed through America’s cities. If blacks seem not to be satisfied with (mostly) uninhibited access to the polls and (close to) proportional representation, this is because they should not be content with these achievements. What really matters in a democracy is getting policies enacted that correspond to people’s views. And on this front, blacks still have a long way to go. Their opinions—on vital issues like crime, welfare, and housing—are too often ignored by elected officials when they conflict with whites’ preferences. This grim reality is well worth getting upset about.

More broadly, the results also highlight other distortions of the political process. Unsurprisingly, in this era of skyrocketing inequality and campaign-finance deregulation, policymakers pay closer attention to the views of the rich than to those of the poor (or of the middle class). More unexpectedly, women are largely ineffectual politically despite their large numbers, high turnout rates, and substantial representation. None of these assets provides them with what a healthy democracy should: laws as reflective of their preferences as of those of men. And Hispanics seem to hold an intermediate position, neither as weak as blacks nor as influential as whites.

So how might the promise of American pluralism be fulfilled—so that public policy is about equally responsive to each group’s views? Unfortunately, this is where we reach the limits of current academic knowledge. We can now quantify the power wielded by different groups. But neither my work, nor anyone else’s, has determined what the causes of group influence might be, let alone how disparities in clout can be corrected.

Still, even in the absence of hard data, it is possible to speculate. In particular, three drivers of group power come readily to mind: participation, resources, and ideology. Perhaps groups whose members engage more actively in politics—by voting but also by attending meetings, contacting representatives, volunteering for campaigns, and so on—have more sway over policy outcomes. Or maybe wealthy groups have more influence. If money is the mother’s milk of politics, as Jesse Unruh once said, it is the affluent who control more of this vital resource. And ideology may matter too. Extreme groups may have trouble making deals and forming alliances as effectively as moderate ones.

These hypotheses point to an agenda of sorts for those who are troubled by the power imbalances of modern American politics. Automatic voter registration and other electoral reforms might broaden political participation. Reducing income inequality, and introducing public financing or more vigorous campaign-finance regulation, could lessen the impact of money on politics. And groups could try aiming for tactical moderation even as they keep fighting to achieve their ideological goals.

Ultimately, it is unclear if these policies would eliminate the power differentials that my study identified. But they would likely help. The first step, though, is simply recognizing the hurdles facing many groups as they struggle to translate their views into law.