Monthly Archives: March 2015

How Donors Distort Democracy

It’s early, but arguably the most important paper of the year has already been released. The author, Michael Jay Barber, finds persuasive evidence that those who donate more than $200 (.22% of the population in 2014), wield more influence over our political system than anyone else.

Demos has explored this issue extensively, arguing that the rise of money in politics has hampered both racial justice and economic mobility. However, the new evidence makes our case even stronger, and more important than ever.

Barber mailed out 20,500 letters to people who donated to Senate campaigns in 2012 and then analyzed how similar the positions taken by politicians were to donor and  non-donors in the state. He notes that politicians almost perfectly align with the views of donors, align strongly with the preferences of those who are from the same party and voted for them, less well with those who voted for them and not at all with their constituents (see chart).

But the paper contains another disturbing fact: donors are polarizing our politics.

As I’ve noted before, polarization is an important driver of inequality, and the polarization is largely driven by elite control of the political system. Barber’s evidence supports this argument: he finds that donors are far more extreme than voters in general and supporters of the candidates (see chart).

He observes that while about 60% of Republican voters dislike the Affordable Care Act, opposition is nearly unanimous among donors. It is donors, not Republican voters, that are driving the numerous efforts to repeal Obamacare. Barber concludes that even among the rich, donors are far better represented than voters, and more extreme.

He finds support in a recent paper by Jesse Rhodes and Brian Schaffner, who wrote (in a study of the House of Representatives), “we estimate that millionaires receive about twice as much representation when they comprise about 5% of the district’s population than the poorest wealth group does when it makes up 50% of the district.” They find that conservative politicians are more responsive to the donor class than liberals (a finding supported elsewhere).

But is the story here that politicians are responding to donors, or donors responding to the positions of politicians. Another paper by Barber suggests that extreme donors are driving extreme politicians. He examines state-level campaign contribution limits and finds that states with higher limits have more polarized legislatures. This suggests that politicians become more extreme to get money from the donor class. Patrick Flavin notes that in states with stricter campaign finance laws devote a large portion of their budgets to programs that benefit the poor. In another study, Walter Stone and Elizabeth Simas find, “that challengers closer to the extreme received greater financial contributions, which enhanced their chances of victory.”

In a 2014 study, Barber, Brandice Canes-Wrone and Sharece Thrower examined how Presidents respond to the donor class. They find that, “there is no clear relationship between general public opinion on an issue and the president’s support for that issue.” However, they find,

when examining presidential issue support alongside co-partisan donor issue support, we observe a clear positive relationship. Particularly, as donors increase their support for an issue, so does the president. While this conforms to our expectations, we also see that there exists a similar positive relationship between the positions of the president and non-donors of his party.

To disentangle whether Presidential candidates were responding to donors or simply partisan supporters, they examined issues where partisan supporters and donors disagreed. In these conflicts, they find,

a one percentage increase in co-partisan donor support on an issue correlates with a one percentage increase in the probability that the president supports that same issue. Further, we find no significant effect of non-donor co-partisan opinion or general public opinion on presidential issue support.

That is, the views of the general public and even people of the same party of Presidential candidates has no effect on what issue a Presidential candidate supports. This is consistent with the work of Martin Gilens who finds,

economic elites and organized groups representing business interests have substantial independent impacts on U.S. government policy, while average citizens and mass-based interest groups have little or no independent influence.

American democracy is controlled by a small donor class who advocate in favor of issues that the vast majority of Americans oppose.

This piece originally appeared on Policyshop.

Are millennials tolerant racists?

Millennials are considered the most diverse, tolerant and racially progressive generation in U.S. history. “The younger generation is more racially tolerant than their elders,” the Pew Research Center declared in a 2010 report based on analysis of more than two decades of data. Commenting on the report, The Chicago Tribune’s Ted Gregory went one step further, arguingmillennials are “the most tolerant generation in history.”

America’s newest generation is more racially progressive than its predecessors’. For example, the Pew study shows that millennials are more likely to support interracial marriage and dating and are generally moreaccepting of immigrants. They see themselves as racially progressive as well. According to a 2014 survey (PDF) of millennials conducted by MTV and David Binder Research, nearly all respondents said they believe “everyone should be treated equally, regardless of race,” 72 percent said “their generation believes in equality more than older people,” and 58 percent believed “racism will become less and less of an issue” as they take on leadership roles. More than half believe that racial bias is “small but real” and “subtler” than it was in the past.

However, such pervasive sentiments do not reflect reality. In fact, beneath the facade of a colorblind generation remains a deep underclass. And millennials are not as racially progressive as the narrative suggests. Studies show that white millennials have opinions similar to older generations’ on issues such as race. A closer look at the Pew Center’s data and other relevant research shows a less-reported but revealing fact: Much of the purported tolerance of the millennial generation is due to the inclusion of more people of color in the pool.

White millennials

Spencer Piston, a professor at the Campbell Institute at Syracuse University, examined the 2012 American National Election Studies racial stereotype battery. Respondents were asked to rate whites, blacks, Hispanics and Asians according to how hard working or intelligent they are. “White millennials appear to be no less prejudiced than the rest of the white population, at least using this data set and this measure of prejudice,” Pistonsaid during a recent New York magazine interview. A 2012 Public Religion Institute poll found that 58 percent of white millennials say discrimination affects whites as much as it affects people of color. Only 39 percent of Hispanic millennials and 24 percent of black millennials agree. Similarly, the MTV poll found that only 39 percent of white millennials believe “white people have more opportunities today than racial minority groups.” By contrast, 65 percent of people of color felt that whites have differential access to jobs and other opportunities. Still, 70 percent of millennials said, “it’s never fair to give preferential treatment to one race over another, regardless of historical inequalities.”

White millennials are more optimistic about the state of race relations. For example, a 2014 Pew survey found that 42 percent of white millennials said “a lot” needs to be done to achieve Martin Luther King Jr.’s dream of racial equity, compared with 54 percent of millennials of color. One-fifth of white millennials said “a little/none at all” needs to be done. There is a significant racial gap in terms of attitudes about how well blacks and whites get along. About 30 percent of nonwhite millennials said whites and blacks don’t get along “too well/not at all well,” compared with 13 percent of white millennials. These gaps remain unchanged across generations. All in all, when the Pew data are disaggregated, they shows large and persistent racial gaps that are obscured when the generations are considered as a whole.

Millennials don’t fare better than their parents on implicit racial bias either. The nonprofit Project Implicit conducts an association test to measure automatic and unconscious preference for European or African faces. A study of 2.5 million voluntary tests taken from 2000 to 2006 found very little variation on implicit bias across age groups, with the exception of those 60 or older. The chart below shows the results of the implicit-association test (IAT) and individuals’ self-reported bias, with a score of 2 indicating a strong bias toward whites and –2 indicating a strong bias toward blacks. The old and young showed differences in their self-evaluation of racial bias, with older people off by 0.38 points and those in the youngest two brackets underreporting their bias by 0.52 on average.


The fact that millennials perceive themselves as uniquely tolerant may make them more likely to practice or accept discriminatory behavior. “A representative panel of Americans interviewed immediately before and after the election [of Barack Obama] reveals a roughly 10 percent decline in perceptions of racial discrimination,” Nicholas A. Valentino and Ted Brader, wrote in a 2011 study in the journal Public Opinion Quarterly.

But the dramatic change in perceptions was clearly symbolic. Valentino and Brader found that “declines in perceived discrimination were associated with increases in negative opinions of blacks and heightened opposition to both affirmative action and immigration.” A large body of research supports this finding. For instance, a 2009 study by Vincent Hutchings found (PDF) “scant evidence of a decline in the racial divide” from 1988 to 2008 on policies that would alleviate racial inequality. Even more startling, Hutchings noted, “younger cohorts of whites are no more racially liberal in 2008 than they were in 1988.”


Millennials are more likely to view Obama’s electoral victory as proof that racial discrimination has been alleviated. Research shows that his election led to what is called symbolic racism, the belief that discrimination no longer exists and that persisting inequalities are due to blacks’ weakness. When whites were reminded of Obama’s victory (regardless of whether they supported him) they were more likely to say that racism is behind us and that blacks receive undeserved advantages. They were more likely to say that a continued push for racial equity is unjustified and that any failure of blacks to succeed is their own responsibility.

A 2009 survey of 74 undergraduates at the University of Washington found that Obama’s election led to a decline in the number of respondents who said there was a need for racial policies such as affirmative action, workplace diversity policies and measures that boost equitable access to health care. Similarly, while liberal undergraduate students at Stanford University wereprimed to recall their support for Obama over a white candidate, they were more likely to support a white job applicant over an equally qualified black applicant.

A flicker of hope

But there are signs of hope for racial progress. In a 2013 study, Tatishe Nteta and Jill Greenlee examined what they call the “Obama generation” — those born from 1982 to 1992. “It appears that the youngest generation of white Americans is leading the way toward a more liberal racial future, [but] the structure of these attitudes compels us to stop short of predicting a more racially liberal America,” they wrote in the journal Political Psychology. French essayist Albert Memmi’s observation on racism explains the authors’ hesitation. He wrote, “There is a strange kind of enigma associated with the problem of racism. No one, or almost no one, wishes to see themselves as racist; still, racism persists, real and tenacious.” That is, while young white Americans are clearly aware of interpersonal racism, they seem unwilling to address structural or implicit biases. It may be that racial progress will occur simply because there are fewer young whites relative to people of color.

Another hopeful development is that Americans, across all ages, are less racially biased than before. Studies show that the last four decades saw a general decline (PDF) of racial prejudice across all generations rather merely the young becoming more racially tolerant. However, both whites and blacks are less likely to attribute racial gaps to discrimination. Instead a growing number choose no explanation at all, suggesting what sociologist Tyler Forman calls “racial apathy.” Yet racial inequities still exist. And the millennial generation is still deeply segregated. Racial gaps in employment opportunity, income, education, incarceration and wealth are eitherstagnant or growing. If millennials remain utterly unaware of racial reality in America, the gaps will only grow deeper.

This piece originally appeared on Al Jazeera.

Millennials Are Less Racially Tolerant Than You Think

However frustrating the current state of race relations in the U.S., there is, according to various pundits and prognosticators, hope for the future: Millennials, they say, are the most tolerant, race-blind generation in human history. And when they grow up and constitute the bulk of the adult U.S. population, many of the problems that have plagued U.S. race-relations for centuries will simply melt away, relics of a less-enlightened past.

It’s a claim that shows up again and again. A 2010 Pew Research report trumpeted that more than two decades of research confirm that “the younger generation is more racially tolerant than their elders.” In the Chicago Tribune, Ted Gregory seized on this to declare millennials “the most tolerant generation in history.” David Burstein, the millennial author of Fast Futuresaid millennials are “more tolerant … than any generation before them.” Hannah Seligson, also a millennial, sounded a similar note in the Daily Beast, writing that research “reveals that we’ve emerged as the most diverse, tolerant, pioneering, educated, and innovative generation in history.” And it’s not just the pundits: A poll from Reason-Rupe shows that in every age bracket, a majority of respondents say that “tolerant” describes millennials “very well.”

Given that race-based gaps pertaining to employment opportunities, income, education, incarceration, and wealth are either persisting or growing, there’s a welcome sense that help is on the way in the form of a more racially enlightened populace.

The problem with these rosy sentiments is that they’re at least partly false. Those who claim that the rise of the millennials will usher in a new age of racial harmony are cherry-picking or misreading statistics. They’re doing so primarily in two ways: by lumping together all millennials when they report survey findings rather than breaking out white millennials views on racial issues, or by focusing narrowly on a small set of questions about explicit racial beliefs that don’t tell the full story. The fact of the matter is that millennials who are white — that is, members of the group that has always had the most regressive racial beliefs, and who will constitute a majority of U.S. voters for at least another couple of decades — are, on key questions involving race, no more open-minded than their parents. The only real difference, in fact, is that they think they are.

When it comes to certain surface-level statistics, it’s true that millennials as a group are more racially progressive than their parents. Pew data show they are more likely to support interracial marriage and dating and are more in favor of immigration. Nearly all agree that “everyone should be treated equally, regardless of their race.”

Dig just a few inches deeper, though, and there’s plenty of fodder for pessimism. Just ask Spencer Piston, an assistant professor of political science at Syracuse University. He examined the 2012 American National Election Studies racial stereotype battery, in which survey respondents are asked to rate whites, African-American, Hispanics, and Asians according to how hard-working or intelligent they are, and found something startling: Younger (under-30) whites are just as likely as older ones to view whites as more intelligent and harder-working than African-Americans (among the older cohort, 64 percent felt this way, and among the younger cohort the number was 61 percent — not a statistically significant difference). “White millennials appear to be no less prejudiced than the rest of the white population,” Piston told Science of Us in an email, “at least using this dataset and this measure of prejudice.”

Asking people racially tinged questions directly can only get you so far, of course. Social scientists have known for a long time that there frequently exists a gap between how people respond to questions and how they really feel — people are swayed by the expectation of how they should answer. A favorite way around this is to measure implicit bias — that is, forms of bias that the holder might not even be aware of and that can manifest themselves in split-second decision-making. In the most common examples of so-called implicit association tests, words or images are briefly flashed, “priming” subjects to respond to subsequent stimuli — if you’re quicker to pair a black face with the word criminal, to take a hypothetical example, you’re exhibiting more implicit bias, and researchers think these effects extend out of the lab into everyday interactions.

If white millennials were, in fact, significantly more racially tolerant than previous generations, it would show up in implicit association tests. And yet they do no better than many of their older counterparts. For example, a study of 2.5 million voluntary IAT tests from between July 2000 and May 2006 shows very little difference across age groups, with the exception of those 60 or older. Other age cutoffs show a similar result: With the exception of the elderly, who do exhibit significantly more racial animosity, there is little generational difference in implicit bias. What does divide old and young is differences in the accuracy of their self-evaluation of racial bias. While older people underestimate their bias by an average of .38 points on a four-point scale, the youngest two brackets under-report their bias by an average of .52 points on average. Younger people, in other words, are simply more deluded about their own beliefs.


None of this has stopped white millennials from congratulating themselves for being so racially progressive, nor has it staunched their racial optimism. Tellingly, nonwhite millennials aren’t quite so optimistic. According to Pew data, when millennials are asked how well they think whites and African-American get along, just 13 percent of whites answer “Not too well/not at all well,” compared to 30 percent of nonwhite millennials. So there are some obvious disconnects here — both between what white millennials see when they look in the mirror and their real-life beliefs, and between how white and nonwhite millennials view the current pace of progress.

It’s true that America is becoming a more racially diverse place — as is frequently pointed out, it is likely that by sometime around 2050, whites will be in the minority. Hopefully, this diversity will bring with it more understanding. But many observers, appealing to stereotypes about millennials that have dubious empirical grounding, are creating a veneer of false progress. Millennials may eventually usher in a more racially enlightened age, but such a shift will require a deeper understanding of race and racism than many white millennials exhibit, rather than the self-congratulatory rhetoric of a postracial society.

This piece originally appeared on New York.

The unbearable whiteness of America’s donor class

The rise of money in U.S. politics has been widely discussed in the wake of two controversial decisions by the Supreme Court: Citizen’s United and McCutcheon v. FEC. However, the discussion has largely been framed along class lines — with the voices of the poor and middle class drowned out by theexplosion of donations and political activity by wealthy elites. While this is true, there is also an important racial element to the rise of big money in politics. For example, a recent report by Demos found that the dominance of money in politics has slowed racial and economic progress (PDF) in the United States.

Political donations are deeply racialized, with more than 90 percent of contributions above $200 made to presidential campaigns and super PACs in 2012 coming from majority-white neighborhoods. A recent analysis by the Brookings Institution found that all of the most politically active U.S. billionaires, with possible exception of eBay founder Pierre Omidyar, who is of Persian-American origin, are white. Similarly, the top 10 Republican and top 10 Democratic donors in 2012 appear to be white, according to Demos. There are scarcely any comprehensive data about the race of donors. A preliminary report on donors in New York City’s 2009 municipal electionsconducted by Public Campaign found that the lowest-tier contributions came from more diverse neighborhoods than the top-tier donations. Contributors from predominantly African-American neighborhoods accounted for 30 percent of donations of $10 or less but only 5 percent of donations over $2,500, according to the report.

For whites, the trend is reversed, with people from white neighborhoods accounting for 78 percent of donations over $2,500 and 38 percent of donations of $10 or less. The lowest-tier contributors were from neighborhoods where people of color make up 73 percent of the population, while those donating more than $2,500 lived in districts where people of color make up only 26 percent. Small donors lived in neighborhoods with a 17.14 percent poverty rate and a median income of $53,790, while the top donors came from communities with a 5.14 percent poverty rate and a median income of $111,170.


Campaign contributions facilitate access, influence legislative agendas and, of course, help get those agendas passed. The widening gap in the level of donations means that people of color will not be represented as well as whites. In their seminal work on political equality in the United States, John Griffin and Brian Newman found that “whites get what they want more often than do Latinos or African-Americans.” This lack of representation is not simply due to class differences. “This is true even beyond the effects of income differences between the groups and even when minorities make up a substantial proportion of a constituency,” according to Griffin and Newman.

In his book “Uneasy Alliances: Race and Party Competition in America” Paul Frymer found that Democrats ignore the preferences of African-American voters, a “captured minority,” in order to win over white voters. Since blacks have chosen not to defect to the Republican Party, they struggle for representation by the Democrats. “All too often the Democratic Party has taken the black vote for granted, and all too often the Republican Party has written it off,” Jack Kemp, then a Republican nominee for vice president,said during a campaign stop in Harlem in 1999. In a study comparing voter preferences on government spending levels to presidential budget proposals from 1974 to 2010, Griffin and Newman found that “presidential requests are much more responsive to the budget preferences of whites and the wealthy.” They found that the gap for African-Americans is smaller when a Democrat is in office than when a Republican is.

This has important effects on legislation, as whites and people of color differ strongly on many issues, particularly those related to class. As the chart below shows, the differences on preferences about redistribution between whites and people of color (14 points) are stronger than the differences along class lines (10 points). A 2012 Washington Post/ABC News poll found that differences on whether the government should spend money on jobs versus reducing deficit are similarly strong along race lines (23 points) compared with class (11 points). On issues relating to race, recent polls show huge racial gaps on the perceptions of major American institutions.


The funding gap between black and white communities also makes it difficult for candidates of color to raise money. Candidates of color raised 47 percent less money than white candidates in all 2006 state legislative races and 64 percent less in the South. This is an important barrier for potential candidates who already face discrimination by voters. The lack of diversity in candidates is problematic, given the importance of descriptive representation (in which politicians share salient qualities such as race with their constituents) for substantive representation where candidates share their constituents’ policy preferences. “Descriptive representation independently improves the relative representation of minorities” and in some cases leads to political equality, according to Griffin and Newman. In a 2012 study in the American Journal of Political Science, Eric Juenke and Roper Preuhs found that candidates of color represent constituents of color better than white candidates do. However, as Jason Zengerle notes in The New Republic, black representatives, particularly in the South, are increasingly in the minority as white Democrats fall to coordinated campaign money from big donors. At all levels, big money politics stifles the representation of people of color.

The lack of representation has had major effects on people of color, as Adam Lioz documents in a series of case studies. For instance, the private prison system has lobbied extensively to maintain the carceral state, which primarily affects men of color. At the same time, the financial system has engaged in blatant racism, aided by lax regulatory policies. According to a study by the Center for Responsible Lending, among borrowers with good credit, people of color received a high interest rate loan three times as frequently as white borrowers. Reports also show racist lending practices by bank employees. “Employees had referred to blacks as ‘mud people’ and to subprime lending as ‘ghetto loans,’” according to a 2009 report by The New York Times. Such acts were possible because of lax regulation because of themassive political power of the financial sector.

There are several ways to fight the big money bias. For one, public financing has proved effective in eliminating the big money bias in the political system. A 2012 study published by The Election Law Journal found (PDF) that New York City’s donor-matching program has shifted the class and racial diversity of donors. It empowered small donors (a pool more diverse than large donors) and increased the absolute number of campaign contributors. Second, to limit the influence of big money in U.S. elections, we need policies that could boost voter turnout.

However, as Griffin and Newman point out, there are limits to the strategy: The benefits of representation are not as strong for blacks and Hispanics as for whites. In other words, African-Americans who vote aren’t that much better represented than those who don’t. This points to the importance of electing people of color to office, which does more to increase the representation of people of color. Still, a public financing program to reduce the influence of money in politics, sane lobbying reformssame-day registration, an end to felony disenfranchisement and increased descriptive representation could go a long way to thwart the influence of big money in our politics.

This piece originally appeared on Al Jazeera.

On Income Inequality: An Interview With Branko Milanovic

Branko Milanovic is a World Bank economist and development specialist. He’s currently a visiting presidential professor at CUNY’s Graduate Center and a senior scholar at the Luxembourg Income Study Center. His book, The Haves and the Have-Nots: A Brief and Idiosyncratic History of Global Inequality, examines—as the title suggests—income inequality. Milanovic and Demos Research Assistant Sean McElwee recently discussed Milanovic’s research and the major shifts within the inequality research field.

Sean McElwee: You’ve been researching inequality for a long time. Inequality as a topic, for a while, was very unpopular. How has the way that inequality is discussed changed in the last decade?

Branko Milanovic: Well, yes, I’ve been researching inequality, income inequality to be precise, for many years. One could even argue that since the late 1970s to early 1980s it was a topic that had no particular appeal to economists. That is, until five to six years ago.

SM: So then, what happened to change this?

Branko Milanovic: What happened is, I think, first and foremost, the recession. Secondly, the mainstream economics became discredited, not because it could not predict or prevent the crisis, but because it was actually saying, until the very last moment, that the crisis cannot and will not happen.

Plus the realization by many people, when the crisis hit and they could no longer borrow as before, that their incomes had not grown in many years, is what really brought up the issue of inequality. Like, “how is it that some people’s incomes have increased a lot and my income stood still?” It was, of course, known by those few who dealt with income distribution before, but it never penetrated much into the economics profession nor into people’s consciousness. Then, with the crisis, things suddenly changed.

SM:  One of the things that I thought was most interesting about your work is this idea of the citizenship rent.

You said something like 60 percent of your income is determined at birth and then 20 additional percent by how rich are your parents. I was curious if we included race and gender in that, what percentage of your income do we know before you even begin your pre-K.

Branko Milanovic:  Global inequality studies were made possible by two events that occurred at about the same time, in the mid-1980s. First, for the first time ever we had income distribution data for most of the countries in the world. Before, you didn’t have household surveys from the Soviet Republics, China and most of Africa. That changed around mid 1980s/late 1980s.

The second thing that changed was, of course, globalization, the inclusion of China and formerly Communist countries in the world market, and thus much greater awareness of the large income gaps between countries and peoples, and of course greater competition.

The graph (for year 2008) shows on the horizontal axis a person’s position in their own country’s income distribution, and on the vertical axis, a person’s position in global income distribution. Thus, the poorest Americans (points 1 or 2 on the horizontal axis have incomes that put them above the 50th percentile worldwide). Note that 12% of the richest Americans belong to the global top 1%.

It turns out that—depending on the year and how detailed your data are—some 50 to 60 percent of income differences between individuals in the world is due simply to the mean income differences between the countries where people live. In other words, if you want to be rich, you’d better be born in a rich country (or emigrate there). You can see that in the figure here, where very poorest people in the United States have an income level which is equal to that of the middle class in China or even upper middle class in India.

So that’s very striking. At least half of your income is determined by where you live, which for most people is where you were born. Then about 20 percent is due to the income level of your parents. So, your citizenship plus your parental background explain around two-thirds or even 70 percent of your income.

Then, obviously, if I had data for gender, race, ethnicity and other things, which are similarly exogenously “given” to an individual, that percentage would go up, perhaps to more than 80 percent.

Sean McElwee: So it’s at least 80 percent, possibly more?

Branko Milanovic:  I am quite confident that if you were to add other things which, I underline, you did nothing to deserve or to be penalized for, you would go, probably, over 80 percent of your income being thus determined.

The lion’s share of that is due to citizenship, or for the individuals like myself who are just residents of rich countries and still get all the benefits, to the residency. This is what I call “the citizenship premium” or “citizenship rent.”

SM: There is more or less a way to deal with inequalities within countries, but how do we deal with inequalities across countries, especially given the way there’s the borders? The fact is, if we’re worried about the rich controlling the U.S. political system, the rich nations, very much control the international political system.

Branko Milanovic: For global equality there’s no mechanism, there are no clear tools, because there is no global government.

Essentially if you think of global inequality, you might say that you have, other than aid, two other tools. One is increase in the growth rate of the poor countries. And this is how, interestingly, you end up in a very curious position, from having started to worry about inequality ending up worrying primarily about economic growth of poor countries.

The second thing is migration. Migration reduces overall inequality on the assumption that when people from poor countries go to rich countries, their incomes go up.

Obviously, migration, while needed to reduce both global poverty and global inequality is not a panacea. Not everything is rosy: first, it could be that the incomes of the people that stay behind are reduced if most skilled people emigrate. This may still reduce global poverty and inequality but it does raise moral issues since we may end up with some countries that would, in the era of globalization, have to disappear: everyone will be better off if they migrated.

And there is a political issue of absorption of migrants in the recipient countries. Still, migration today, despite much attention it receives, has been relatively stagnant and small. In the last quarter of century, the percentage of people who live in the countries where they were not born has been stable at between 2 and 3 percent. Or to go back to your previous question, for 97% of the people in the world, half of their income is decided at the moment when they are born.

SM: Catherine Rampell in her New York Times review of your book, noted income mobility, and I think it gets at this distinction you like to make between good inequality and bad inequality. 

Branko Milanovic:  The bad or undeserved inequality would be the one that arises from the factors over which you have no control: what was the income level of your parents, whether you were born male or female, what is your race, and things like that.

The good inequality is inequality of effort, work, luck and so on.

SM: Okay. So let’s talk about your paper. There’s been a lot of recent research on GDP growth and inequality and it’s actually kind of shifted almost to the left, it sounds like, with more inequality leading to lower GDP growth. You find that actually it is a little more nuanced than that.

Branko Milanovic: I think that with the paper you mention, which I did in collaboration with Roy van der Weide from the World Bank, the novelty is that we “unpack” both growth and inequality.

In the past, the use of these two aggregate and “unnuanced” measures like GDP per capita and Gini coefficient, regressed on each other across countries, yielded the results that were all over the place: from a positive relationship between high inequality leading to higher growth to a negative relationship, to a very weak or non-existent relationship.

We thought of unpacking both growth and inequality. What does it mean? We don’t look at the growth of the mean only; we look at the growth rate at different points of the income distribution, we look at the growth rate of the very poor, say bottom 20 percent of people, we look at the rate of growth of the median, or among the top 10 percent or 5 percent, or top 1 percent.

Then, similarly, [we] disaggregate overall inequality into inequality among the poor, among the rich.

I think our most interesting finding—based on the U.S. data from 1960 to 2010, huge micro-censuses, conducted once every ten years and which include one percent of the U.S. households—is that if you look at income growth of the bottom 20% or bottom 40% of the population in period  “t+1”, that income growth is negatively correlated with inequality which existed in period “t”.

Let me put it in simple terms. Let’s suppose I’m a guy who is poor. I live in Massachusetts in 1990. Then the rate of growth of my income between 1990 and  2000, would be negatively correlated with income inequality that existed in Massachusetts in 1990. So we do it across all fifty U.S. states. Obviously, overall growth rates have declined over the last 50 years; also the shape of the growth rates across various income levels (the poor, the middle, the rich) has changed, as you can see here in the graph, but the negative correlation between inequality and subsequent real income growth of the poor remains.

It used to be that the U.S. growth was pro-poor, in the sense, that the growth rates among the poor were higher than amongst the rich. Now it’s the opposite. But that particular relationship between inequality and growth of the poor, we find it throughout the entire period. We can only speculate what are the reasons that make inequality be so bad, pernicious even, for the growth at low levels of income.


Now, when you move to the top of income distribution, the story changes. Inequality may not be bad for the rich; actually, it’s positively correlated with their growth rates. So if you’re really a rich person in Massachusetts in 1990, then your growth rate, between 1990 and 2000 is going to be positively correlated with inequality which existed in Massachusetts in 1990.

In other words, the bottom-line is that we believe is evidence there to show that the rate of growth of the poor people is negatively affected by high inequality.

SM: The argument by many people who are center-left has been, “Oh, we now know that inequality is bad for growth, so the rich should get behind measures to reduce inequality,” but if you’re correct they actually do not have that political incentive.

Branko Milanovic: That’s a very good point, that’s your point, we don’t make it in the paper, but it’s very true. If you take what we find in the paper, that the growth coefficient on inequality for the rich is positive, then they don’t have an incentive to fight inequality. For their growth rate, inequality is good. It undermines the case for a sort of self-interest of the rich to be more accommodating.

This interview originally appeared on Policyshop.

No, The Decline in Marriage Did Not Increase Inequality

Co-written with Marshall I. Steinbaum.

In a “letter from the editor” last week, The New York Times’ David Leonhardt claimed that liberals overlook evidence that changing household structure, meaning the relative decline of households headed by a married couple, increases inequality. He wrote, “I’d say that family structure is an area where many liberals are putting more weight on their preconceptions (inequality is bad for society) than on the evidence (changes in family structure are both an effect and a cause of inequality).” But there is substantial evidence that household structure is what’s known as an endogenous variable: People make marriage, divorce, and child-rearing decisions due to their economic circumstances. Furthermore, the causes of rising inequality are the same reasons why household structure has changed: stagnant wages, labor market detachment, and job lock.

When questioned on Twitter about the evidence for his claim that changes in family structure cause inequality to rise, Leonhardt cited a paper by Professor Molly Martin of Penn State. But Martin argues nothing of the sort.

Her paper, “Family structure and income inequality in families with children, 1976 to 2000,” essentially asks1 this question: “To what extent is rising inequality due to growing inequality between types of households (e.g., the gap between married couples and single mothers has grown, and a larger share of households belong to the more disadvantaged group), and to what extent is it attributable to growth of inequality within types of households?” Martin finds that 41 percent of the increase in household inequality from 1976-2000 is attributable to changing family structure.

But in a section entitled “Limitations,” Martin writes, “First, the results do not document causal relationships. I cannot determine the degree to which family structure changes caused the observed changes in inequality.” And Martin concludes an appendix further addressing these issues with “family formation probably reacts to prevailing economic conditions and, in that response, sets the conditions for perpetuating broader inequality patterns.” A single clause from the abstract of her paper, “family structure shifts explain 41% of the increase in inequality,” has been incorrectly cited as evidence that the decline in married-parent households caused inequality to increase, when the author herself disavowed that interpretation both in the paper and thereafter. She notes that, “the relationship between family formation behavior and inequality appears to be declining over time” and even during the period where it was most influential, it accounted for very little of the change. See her chart:

There are further problems with using Martin’s paper as the basis for the argument that changing family structure caused inequality to rise. She uses survey data that top-code the highest income households. That means that households earning more than some threshold are reported as earning at the threshold in order to guard their confidentiality. The Current Population Survey is considered to be informative about the bottom 98 percent of the income distribution (more or less), but the top is where all the action has happened, so to speak. Thomas Piketty and his coauthors have shown the rise of top income shares in numerous papers, and it is why they have developed the World Top Incomes Database to track the incomes of the richest.

In a recent summary of the literature, Bruce Western writes, “Most of the increase in family income inequality was due to increasing within group inequality that was widely shared across family types and levels of schooling.” He finds that, “Though family structure, more than the educational inequality in earnings, is closely associated with the rise in inequality from 1975 to 1995, both effects were small after 1995.”

There is a large literature that seeks to treat changing family structure as an outcome in need of explanation, exactly the sort of analysis that Martin stipulates her paper does not provide. Daniel Schneider and Adam Reichinvestigate whether young men who join unions are more likely to get married thereafter, and find strong evidence that they do in high-quality longitudinal data (meaning that the same individuals are followed over time). Importantly, the effect of joining a union on first marriage disappears when the authors also consider the increase in wages and job security that come with union membership. The implication is that the decline in overall marriage rates and the decline in unionization rates are linked. The reason why is that young men increasingly lack access to secure, well-paying jobs.

In a similar vein, Jennifer Lundquist finds that the military fosters high rates of marriage and low divorce, and notably those outcomes are indistinguishable for whites and blacks in the military, whereas in the larger society black households are significantly more likely to be headed by a single parent. Over a series of papers, she reports on the reasons why household structure is so much more traditional in the military: that employment is stable and that a safety net exists for raising children.

Consistent with the idea that economic wellbeing enables a stable marriage, Yu Xie, James Raymo, Kimberly Goyette, and Arland Thornton construct an individual-specific measure of “economic potential” designed to capture the outlook for young people that may not be discernible from current income or a crude predictor of future income, such as education credentials. They find that economic potential increases the marriage rate for young men, but not for women. On the other hand, potential does not affect the probability of cohabitation for either gender.

In a way, the idea that household formation decisions are an economic outcome should not be surprising. One of the foundational publications of modern economics, An Essay on the Principle of Population by Thomas Malthus, argues that the reason why the birthrate increases in prosperous times is that people have the wherewithal to marry earlier. Malthus argued that this effect eventually drives the standard of living back to equilibrium, which thankfully did not turn out to be true for the most part after the 18thcentury. But the bare point, that stable marriage is an aspiration that falls victim when economic opportunities worsen has been a notable feature of social science scholarship since, including the books When Work Disappears by William Julius Wilson and the recent, and excellent, Labor’s Love Lost by Andrew Cherlin.

Cherlin brings a historical perspective to bear: the last time marriage rates for working Americans were low and age at first marriage was high was during the Gilded Age in the late 19thcentury, when inequality was also high. He attributes the rise and fall of the working class family to the rise and fall of the manufacturing sector and stable employment opportunities for young men. He also points out that the working class family was a gendered institution, that it was premised on a male breadwinner earning enough to support a female home-maker, and that its decline is in parallel to the relative expansion of women’s opportunities for employment outside the home. Thus, we can see the prevalence of married households as reflecting both the absolute economic opportunity of men and women and the relative economic opportunity of women. On the latter point, Judith Hellerstein and Melinda Morrill find that divorce rates fall when unemployment is high, suggesting that when economic outcomes outside marriage are worse, spouses remain in marriages they would otherwise have left.

It is facile to divide rising inequality into “between” and “within” effects with respect to household types, and to argue that since inequality between types has grown and more households are now in worse-off types, changing family structure has caused inequality to increase. The evidence shows that family structure has changed because economic opportunities for most people have worsened. Why has that happened? There are some suggestive answers, but much more research is necessary. Leonhardt’s claim that changing family structure causes rising inequality simply doesn’t hold up.


The paper decomposes the change in a measure of inequality called the Mean Logarithmic Deviation as calculated from the March Current Population Survey into “between” and “within” effects.

Democrats Finally Found a Smart Way to Stop Wall Street’s Reckless Behavior

This piece was co-written with Lenore Palladino.

For too long, it appeared that many Democrats were trying to fight economic inequality with policies like the minimum wage while ignoring the 800-pound gorilla, Wall Street. But Rep. Chris Van Hollen on Monday unveiled legislation to cut taxes for those earning less than $200,000, while partially paying for the proposal with a financial transaction tax (FTT), projected to raise $1.2 trillion over the next decade. With this proposal, Van Hollen is recognizing that without reducing financial speculation, it’s impossible to address inequality or to leave Wall Street’s risky practices in the past.

The small FTT in this billwhich also includes provisions to boost stagnant wages and close lucrative tax loopholeswouldn’t burden longer-term investors. The tax is applied to every transactionthe sale and purchase of a stock, bond, or other financial instrumentso as long as the investor holds the investment for a decent period of time, the tax is a tiny percentage of their overall portfolio and won’t drastically alter their trading behavior. It’s the high-frequency traders who have fought this tax tooth and nail, and who will gear up to fight it now, because if you trade multiple times a millisecond then your tax burden will be higher.

High-frequency trading creates systemic risk. Taxing it would reduce the incentive for the financial sector to chase new bubbles, driving out “noise traders” who make markets more volatile without improving capital intermediation (the purpose of the financial system). This is the argument of economists including John Maynard Keynes, Lawrence Summers, Victoria P. Summers,  and Joseph Stiglitz: that reducing the “whirlpools of speculation” is one of the best method for risk reduction. We’ve seen the results of volatility among such traders in the flash crashes, where huge amounts of speculative trading can crash very, very quickly.

Given the recent experience of Dodd-Frank (not to mention the mounting attacks on it), and the failure of regulators in the lead-up to the crisis to accurately understand what was occurring and put the brakes on it, market-based approaches like a FTT are crucial to making sure Wall Street speculation doesn’t bring down the economy again. There’s also lots of evidence that all of this increased trading hasn’t actually made the financial sector more efficient, and that it has been a main driver of economic inequality. Because stock ownership is concentrated at the top (the bottom 90 percent own only 9 percent of stocks and mutual funds), the benefits of finance have accrued to the one percent. A 2011 studyfinds that “financialization accounts for more than half of the decline in labor’s share of income, 10 percent of the growth in officers’ share of compensation, and 15 percent of the growth in earnings dispersion between 1970 and 2008.” Thomas Philiponestimates that inefficiencies in the financial sector cost the U.S. economy $280 billion every year. Other studies support the conclusion that much of the sector subsists on rent-seeking.

Financial transaction taxes are nothing new. New York State, for example, had one in place from 1906 to 1981 (and one remainslegally on the books, but it is refunded to the trader). Out of favor in the 1980’s and 1990s as Wall Street dominated the political and economic agenda, they have gained increasing interest amongpolicymakers. And they’re not just on Occupy Wall Street’s agenda: Nobel laureates like Stiglitz and Larry Summers have called for modest transaction taxes to promote stability and raise revenue, and even the International Monetary Fund has cautiously supported a financial transaction tax. From 1914 to 1966, the United States levied a 0.02 percent tax on sales and transfers of stock. To this day, the SEC is funded by an incredibly small FTT. More recently, Senator Tom Harkin and Representative Peter DeFazioproposed the “Wall Street Trading and Speculators Tax Act,” which would assess a tax of 0.03 percent on trades of stocks, bonds, futures, options, swaps and credit-default swaps, and would generate $352 billion over 10 years. Europe is moving forward with their own FTT.

For the last two decades, taking on Wall Street hasn’t always been a focus for the Democrats. President Ronald Reagan initiated the rise of finance, but President Bill Clinton certainly helped with the passage of Gramm-Leach-Bliley Act, which dismantled key New Deal regulations (including large swaths of Glass-Steagall). The massive amounts of money finance sprinkled on both parties made them virtually immune to regulation. Finance, Insurance and Real Estate (FIRE) companies have dramatically increased their spending on lobbying in recent years, from $609,523,625 inflation-adjusted in the 2000 cycle to $996,406,725 in the 2012 cycle. Recently, IMF researchers linked lobbying to the rise of subprime mortgages and the opaque securitization schemes that left the financial crisis on a precipice.

As Nomi Prins has documented, financial power players have been deeply intertwined with the political process for a long time. However, finance’s political heft has never been greater, to the extent that a recent deregulation amendment was essentially written by financial industry lobbyists. In the wake of Dodd-Frank,bankers were granted 14 times more meetings with the Commodities Futures Trading Commission, the Treasury, and the Federal Reserve Board than pro-reform groups. Political scientist Christopher Wiko notes, “as the Democratic Party coalition began to include more professionals and managers, and as unions declined, the negative relationship between Democratic control of government and financialization weakened.” That is, many Democrats stopped worrying and learned to love the financial sector.

Van Hollen’s proposal is an attempt to reverse this pattern. It’s well past time to get out of these whirlpools of speculation and onto stable, dry land.

This piece originally appeared on The New Republic

Why Obama’s Tax Plan Matters

Today, Obama will deliver the State of the Union address, which will focus on inequality, which he has previously called, “the defining issue of our time,” and which was recentlyhighlighted by a proposal by Chris Van Hollen. In anticipation of his speech, he has put forward a proposal to increase taxes on the wealthy and cut taxes for the middle class. The proposal is welcome.

Obama’s plan is broadly aimed at wealth inequality and closing tax loopholes that allow the rich to pay less in taxes than many middle class households. His plan includes a hike the capital gains tax rate (to the level prevailing during the Reagan administration), closing the trust fund loophole that allows the wealthy to avoid inheritance taxes and a modest fee on banks (who benefit hugely from too-big-to-fail).

While the right often complains that many Americans don’t pay taxes, symbolized by Romney’s famous 47% comment, these claims are absurd. For one, Romney was only talking about taxes at the federal level, and a recent study from the Institute on Taxation and Economic Policy shows that state and local taxes are heavily regressive. Across all states, the poorest 20% pay 10.9% of their income in state and local taxes compared to 5.4% for the richest 1%.

At the same time, across the developed world, tax rates have become less progressive, as illustrated by Thomas Piketty and Emmanuel Saez. In the US, estate taxes and corporate taxes have declined dramatically, while payroll taxes (which affect the poor) have increased dramatically.

Worse still, the porous federal tax system allows the rich to take home huge tax breaks. A recent CBO finds that the largest 10 tax breaks primarily benefit the rich. Recently, Al Jazeera reported that in 2010 the richest 400 Americans paid only 18% of their income in federal taxes, a dramatic decline from the 29.9% they paid in 1995. Therichest 400 households took home a full 16% of all capital gains in 2010—they make up .00025% of the population.

The rich receive the most in capital gains because they own nearly all the assets, and for this reason, Obama’s choice to target capital gains and large holdings of wealth is smart.

Much of the rise in inequality has come from rent-seeking and inheritance, rather than innovation. By singling out unproductive capital for taxation, he forces those who support inequality to defend an idle wealthy class.

But Obama will struggle to pass any legislation increasing taxes. As Demos showed inStacked Deck, the donor class isn’t interested in higher tax rates. Sadly, there is strong evidence that the donor class has an outsized influence on politics, through both access, agenda-setting and higher voter turnout. A 2013 study in The Journal of Economic Perspectives finds (PDF):

In 1980, the top contributor … gave $1.72 million (in 2012) dollars, nearly six times the amount given by the next largest contributor. In 2012, the two largest donors were Sheldon and Miriam Adelson, who gave $56.8 million and $46.6 million, respectively. Other members of the Forbes 400 accompany the Adelsons; 388 current members are on record as having made political contributions. They account for 40 of the 155 individuals who contributed $1 million or more to state and federal elections during the 2012 election cycle.

The rise in wealth has coincided with an increase in political power – which will make taxing the rich more difficult. However, it is increasingly necessary – 80 people in the entire world have the same wealth as the bottom 50% (some 3 billion people) combined.

The usual suspects will likely be rolled out in defense of low tax rates – conservative commentators frequently claim that low taxes increase economic growth. These arguments are facile and a large literature shows them to specious. In fact, there are reasons to believe that increasing taxes on the rich to fund public investment and middle class growth could actually boost the economy. More importantly, these arguments are being disproven before our eyes: Kansas, which was supposed to be the conservative model, massively cut taxes and the state coffers are hemorrhaging cash. The governor is now proposing new taxes to make up for falling revenues.

Obama’s proposal is bold and necessary. The problem is how to get it passed a government increasingly bought by the 1%.

This piece originally appeared on Policyshop.

The Income Gap at the Polls

In 1986, the economist John Kenneth Galbraith declared, “If everybody in this country voted, the Democrats would be in for the next 100 years.” But for decades, the consensus among scholars and journalists has been the opposite. In their seminal 1980 study on the question, using data from 1972, political scientists Raymond Wolfinger and Steven Rosenstone argued that “voters are virtually a carbon copy of the citizen population.” In 1999, Wolfinger and his colleague Benjamin Highton again came to the same conclusion: “Outcomes would not change if everyone voted.” Their argument rested upon the fact that polling data did not show large differences in opinions on most issues between those who voted and those who did not.

However, a growing literature both within the United States and internationally suggests that, in fact, policy would change rather dramatically if everyone voted.

Does this mean that Galbraith was right all along? Not exactly. The reason for the recent shift in the findings is not that the early studies were wrong, but that the preferences of voters and nonvoters are becoming increasingly divergent. In a paper published in 2007 and later expanded into a 2013 book, Who Votes Now, political scientists Jan Leighley and Jonathan Nagler found that wide gaps between voters and nonvoters have opened up when it comes to class-based issues. They argued further that the seeds of these differences were apparent in earlier data, but Wolfinger and Rosenstone overlooked the gaps by focusing on broad ideological labels (liberal or conservative) rather than specific policies. Voters, Leighley and Nagler found, are more economically conservative; whereas non-voters favor more robust unions and more government spending on things like health insurance and public schools.

Other data collected on the national and state level support Leighley and Nagler’s thesis.  A 2012 Pew survey found that likely voters were split 47 percent to 47 percent between Obama and Romney while non-voters preferred Obama 59 percent to 24 percent, a 35 point margin. A 2006 Public Policy Institute of California (PPIC) study found that non-voters were more likely to support higher taxes and more government-funded services. They were also more likely to oppose Proposition 13 (a constitutional amendment which limits property taxes), dislike thenGov. Arnold Schwarzenegger and support affordable housing.

It so happens that the gap between voters and non-voters breaks down strongly along class lines. In the 2012 election, 80.2 percent of those making more than $150,000 voted, while only 46.9 percent of those making less than $10,000 voted. This “class bias,” is so strong that in the three elections  (2008, 2010 and 2012) I examined, there was only one instance of a poorer income bracket turning out at a higher rate than the bracket above them. (In the 2012 election, those making less than $10,000 were slightly more likely to vote than those making between $10,000 and $14,999.) On average, each bracket turned out to vote at a rate 3.7 percentage points higher than the bracket below it.

This class bias is a persistent feature of American voting: A study of 40 years of state-level data finds no instance in which there was not a class bias in the electorate favoring the rich—in other words, no instance in which poorer people in general turned out in higher rates than the rich. That being said, class bias has increased since 1988, just as wide gaps have opened up between the opinions of non-voters and those of voters.

Recent research tells us that this voting disparity—in class and in opinion—has tremendous impact on policy. State-level research suggests that higher voter turnout among the poor leads to higher welfare spending. A 2013 study found that turnout inequality directly predicts minimum wages, children’s health insurance spending and anti-predatory lending policies. And studies at the state level have found that a higher class bias in the electorate actually leads to higher levels of income inequality.

Interestingly, this doesn’t just play out on the national level. Vincent Mahler of Loyola University, who has studied international voter turnout extensively, determined that both voter turnout and inequality of turnout are strongly correlated with income redistribution (see charts).

His most recent study of voting in 14 countries, conducted with political scientists David K. Jesuit and Piotr Paradowski, found that increased turnout among the rich leads to less redistribution, while increased turnout among the poor increases redistribution. This confirms his previous two papers, which together draw on nearly four decades of data across more than a dozen countries. As both of his charts show, the United States is unique among developed countries in turnout inequality, and it has the second lowest voter turnout of all OECD countries. The chart below shows the turnout differences between the richest quintile and the poorest quintile in the United States. The average gap across the countries studied is 8.4 percentage points, while the gap in the United States is nearly three times that—23.6 percentage points.

In a fascinating paper, political scientist Henning Finseraas calculated the class bias and “anti-redistribution bias” of 13 countries. (A higher class bias indicates higher voting levels among high-income voters, while a higher anti-redistribution bias indicates that those who vote are more opposed to redistribution than those who don’t.) Finseraas found that, among the countries he studied, the United States tops the charts in both measures and that, across all countries studied, those who prefer redistribution are less likely to vote than those who do not.

Why does the United States have such depressed voter turnout among the poor? For one, the United States has numerous barriers to voting that don’t exist elsewhere. Nearly 3.5 million felons were barred from voting in 2014 due to felony disenfranchisement, which exists in no other country Mahler studied and which has influenced the result of Senate and presidential elections (including possibly the 2014 election). A number of countrieshave compulsory voting laws which boost turnout. Most European countries that require IDs to be shown at the polls have national ID cards that are provided for free to all citizens. The United States is also an outlier in that it doesn’t have automatic voting registration. In most countries voter registration is compulsory and therefore either universal or almost universal. The United States is also rare among developed countries in the level of politicization that occurs around voting rights.

Exit data from the 2014 midterms indicate that the class bias of the electorate remains strong. Those making under $50,000, who account for 48 percent of the population, made up only 36 percent of voters, while those making over $100,000 made up 30 percent of voters, but only 22 percent of the population. Given the almost unprecedentedvoter suppression in the wake of Shelby County v. Holder, this is not entirely surprising. Voter suppression measures disproportionately affect the poor and people of color, who do not have voter ID, struggle to get time off work and are less likely to be registered in the first place. At the same time new barriers were created, measures that are proven to decrease the class bias of the electorate, like Same-Day Registration were rolled back.

Numerous factors have been shown to contribute to the recent rise in economic inequality. For one, there’s the increasing influence of money over our political system. In 1980, the top contributor gave $1.72 million (inflation adjusted) in campaign contributions, while in 2012 Sheldon Adelson gave $58.8 million and his wife $46.6 million—more than the residents of 12 states combined. There’s also the fact that the U.S. has a majoritariansystem rather than a parliamentary one. But the evidence also suggests that wide gaps in turnout also have a significant effect on inequality, and we shouldn’t ignore it. Reducing these voting gaps and increasing democratic participation in the economy will force policymakers to consider the interests of the working-class.

This article originally appeared on Politico

The real reasons Democrats can’t win

In the run-up to the 2016 election, Republicans are trying to position themselves as the party of the middle class. In a recent essay, Thomas Edsall writes, “The Republican appropriation of leftist populist rhetoric (and even policies) poses a significant threat to liberal prospects in 2016.” It may well work, but not because Republicans are in fact reformist, but rather because voters and pundits eschew data and instead focus on rhetoric. When it comes to actual empirical evidence, the answer is indisputable: Democrats preside over far more income growth for the middle class than Republicans.

Princeton University’s Larry Bartels has two studies on politics and income distribution, and together they encompass almost a century. His finding: under Republicans, the poor and middle class see almost no income growth, while under Democrats, they see dramatic growth (see charts). As he notes elsewhere, even after numerous controls, these partisan differences remain. “Every Republican president in the past 60 years has presided over increasing income inequality, including Dwight Eisenhower in the midst of the ‘Great Compression’ of the post-war decades,” Bartels writes. “And every Democratic president except one (Jimmy Carter) has presided over decreasing or stable inequality.”

In another recent study, Alan Blinder and Mark Watson find that on a number of economic indicators, the country fares far better when a Democrat is in office. GDP growth is 1.8 point higher under a Democratic presidency, unemployment is lower, corporate profits are higher, the S&P grows faster and wages grow faster. This difference is not found in other countries, suggesting that the particularly rabid nature of American conservatism may be an important factor. It could also be that the effect is purely luck (although there is evidence to suggest that left-wing governments can facilitate growth). But the fact that the economy grows faster under Democrats is not enough to explain why the middle-class fares better. As the chart below shows, much of the distribution leg-work occurs after taxes and transfers. This isn’t to say Democrats don’t shape the pre-tax distribution (they do), but rather that simple differences in market distributions of income can’t explain the difference.

As John B. Judis argued — contrary to his seminal proposition of an “Emerging Democratic Majority”  — the future now belongs to the Republican party. It’s increasingly likely that Democrats will continue to have a slight advantage in the electoral college, but struggle elsewhere, for reasons I’ve previously discussed. So, while Judis’ thesis that middle-class whites are dramatically shifting right is contestable, he raises an important point: Middle-class Americans like services but dislike taxes, and Democrats currently appear to be the party of taxes. And so, the struggle for Democrats is what Suzanne Mettler refers to as the “submerged state.” That is, the way the government actually benefits the middle class often goes unseen, while taxes, particularly the income tax, are very obvious. Mettler notes that our federal tax code is full of handouts like the Mortgage Interest Deduction, but these tax benefits primarily benefit the affluent and middle class. “Our government is integrally intertwined with everyday life from healthcare to housing, but in forms that often elude our vision,” she argues.

The implication is that many people who believe themselves independent of government support in fact rely heavily on it. The Congressional Budget Office estimates that the 10 largest tax breaks cost the government $900 billion in 2013. But the benefits accrue to the wealthy: The top 1 percent gets 17 percent of the benefits and the bottom quintile only 8 percent. As the New York Times reported in 2012,

“The government safety net was created to keep Americans from abject poverty, but the poorest households no longer receive a majority of government benefits… The share of benefits flowing to the least affluent households, the bottom fifth, has declined from 54 percent in 1979 to 36 percent in 2007, according to a Congressional Budget Office analysis published last years.”

As Christopher Howard notes in his book, “The Hidden Welfare State,” “There is, still, a misconception that U.S. social programs primarily benefit the poor. That is not true for the visible welfare state direct expenditures, and it is an absurd claim to make about the hidden welfare state.”

As the political science literature shows persuasively, Democrats are far better for economic growth, and particularly middle-class and poor income growth, than Republicans. Yet even liberal commentators often fail to notice this. (Kevin Drum, for example, argues that “Democrats simply don’t consistently support concrete policies that help the broad working and middle classes.”) By focusing on major policies, these critics miss what Nathan Kelly calls, “market conditioning,” or the ways in which left-leaning governments shift market distributions through regulation, monetary and fiscal policy and other non-explicitly redistributive functions. In fact, there is a strong literature showing that parties on the left shift the income distribution. One notable example: While conservatives savagely attack unions, which dramatically shift the income distribution, while Democrats leave them alone. Further, Democrats tend to favor expansionary economic policies, while Republicans try to clamp down on inflation — which primarily benefits the rich. A cross-country study by Isa Camyar finds that firms perform better under left-wing governments because such governments spend more money on public fixed investment. This will naturally lead to higher wages. And while the real value of the minimum wage has increased 16 cents a year under Democratic presidents, it has decreased by 6 cents per year under Republicans. Liberal governments also do more to reduce unemployment, which is significant for earnings at the bottom of the income distribution.

So, while there is evidence that the in the era of globalization and finance, liberal parties can’t do as much to impact the distribution of income, it is nonetheless clear that liberals matter, as Bartel’s recent data extend to 2012.

On the other hand, the conservative premise — that inequality will increase growth and thereby benefit the poor and middle class — has been so thoroughly demolished that it can’t be stated with a straight face. A large body of empirical literature suggests that massive tax cuts for top earners do little but increase incomes at the top. Branko Milanovic and Roy van der Weide find, “high levels of inequality reduce the income growth of the poor and, if anything, help the growth of the rich.” Dan Andrew, Christopher Jencks and Andrew Leigh find that whatever modest effect that inequality has on growth is mitigated by the impact of inequality on the bottom. Intelligent conservative commentators have essentially surrendered the supply-side debate.

It’s clear a Democratic Party is better for middle- and low-income growth. However, while it’s entirely mythological that the poor tend to vote Republican, it is still true that Democrats have trouble with the white middle class. An important reason for this is that Democrats are often seen as the party that benefits the poor, particularly poor black Americans. My investigation of ANES data shows this phenomenon in action: Whites are opposed to welfare, but support helping the poor (see chart). A large literature shows that opposition to welfare and government is driven by racial animus. To this day, the historical portion of slaves in a county predicts Republican support and racial resentment.

Because of our strange political system, middle-class white Americans can therefore believe that the government only takes from them and only helps black Americans. In fact, government programs frequently exclude people of color, and those that do benefit them are always on the chopping block. Whites, particularly Southern whites, often oppose programs that would help them simply to ensure that people of color remain in abject poverty. They see the bad parts of government in the form of taxes, but their welfare is hidden in a maze of subsidies. If you had to collect the mortgage interest deduction at the welfare office, Democrats would never lose another election.

This piece originally appeared on Salon