My politically conservative father has not voted for a Democrat since Jimmy Carter. This is a message that appeals to him. CEO pay & decreasing living standards in his area (very white, very rural): two things he talks about, constantly.
If the message is racialized, it becomes less salient. To his credit, he is fair-minded enough to acknowledge our history of racial inequity and how it factors into economic outcomes. Still, these conversations are meaningless for his community unless they are tied to economics, because they only have poor people that are white. Southwest Virginia not likely to go progressive any time soon, but I think this should be the mantle of the oft-malpracticed Democratic strategy to win in rural communities.
Some of your points relate to the narrow group of billionaires, while the top quarter probably includes most of your white audience. The slippage in who is "rich" is a big problem in framing solutions about class. I have a feeling the necessary solutions involve redistribution from the top quarter, but in the comments you are telling people they can ignore that point if they don't like it!
This question is raised (and purportedly answered) all the time but I haven't seen a lot of data to back it up, even though it's really important! I'd be very curious to see clear quantitative assessments of how much money you'd need to raise to meaningfully raise living standards of the bottom 50%, and whether you'd really need to tap the richest quarter for more taxes or just the top 10% or 1%!
Of course even as I write this I realize attempting to assess this in the objective way I describe is impossible in a world where interest rates are zero or even negative, so the proper compromise is to simply spend the money without raising taxes on anyone... The proposition of raising taxes for the *purpose* of addressing inequality *by taking money from rich people* is not an objective quantitative question.
I agree with the political argument that it’s better for progressive coalitions to focus on economic redistribution and maybe public policy more broadly in a race neutral way. I also agree that better enforcement of laws that currently exist can go a long way in ameliorating discrimination.
That said, I don’t think there’s enough acknowledgement of how much US political economy is racialized, and how challenging it is to execute your strategy.
For example, Obama tried to follow this logic pretty closely, and we were barely were able to get Obamacare passed and maintained (to this day). And the policy impact of Obamacare is less likely to reach Blacker states like Alabama vs. whiter states like Massachusetts.
And this isn’t to say that Obamacare wasn’t the best that we could do. It’s suggesting that those design choices reflect a fairly racialized political economy even when you’re trying to do the opposite, which has the intended effects.
So the politics that we’re seeing is an effort to change the country’s political economy by not pretending that race isn’t a considerable animating factor for the behavior of political actors, but calling it out, and rejecting it as the only way to get out of our toxic political dynamics. It’s just not sustainable for the people living through it.
To be a bit clearer, what’s missing from this argument that makes it much weaker is that if you do this analysis against a whole range of socioeconomic outcomes around health, education, criminal justice, civic engagement, social capital, etc. there are huge racial gaps across class there as well. And living under those conditions is disempowering and miserable.
For example, economic mobility for middle class black boys is lower than for poor white boys! Rich black boys are more likely to end up in jail than poor white boys! That’s not a class problem. And the experience of living in that hierarchy isn’t properly reflected in the outcomes that we’re focused on.
An important question to consider, is how do you build a politics that doesn’t acknowledge that? Barack is right! Defund the police is a terrible slogan - but I am not sure “Reform the police” would have created better politics at the margins. It would be demagogued in the same way, and not forced a reckoning with the issues we need to have.
I am halfway convinced that your approach would work, but can we go to the places in the country that are closer to this approach vs. not and are their outcomes any better or any less racialized?
Most people of color don’t want to have every argument and analysis on the basis of race, and many are aware that wealthy POC can learn to live comfortably in an unsustainable class hierarchy. They also know the impact of race permeates everything and it has to be extricated in parallel with executing a more class based politics and policy.
Lots of excellent points, Nick. And to your point about racialized political economy, I'm surprised that Matt is not familiar with the arguments of many social scientists that a major reason for low support for a European-style welfare state in the USA is racial diversity: people are less likely to support redistribution if they view it as benefiting undeserving minorities. Trying to sell your programs in purely economic terms doesn't necessarily make the racial dynamics disappear.
An example of the argument that racial animus in the US reduces support for welfare-state programs, from Alesina Glaeser and Sacerdote: https://www.nber.org/papers/w8524
But the traditional Democratic approach - e.g. as recently as Obama, as Matt points out - has been to respond by downplaying race. The new Democratic approach of *highlighting* race seems likely to aggravate the dynamic described by Alesina et al.
Maybe within the college environment the direct approach to tackling racial animus seems far easier than in society at large?
A similar argument from a Canadian perspective: Joseph Heath suggests that it's human nature to divide into groups, but not necessarily by *race*. So in a multiracial society, you want to downplay the salience of race as much as possible. (For example, the coalitions represented by the major Canadian political parties are not divided by race.) https://induecourse.utoronto.ca/on-racism-and-race-consciousness/
I think Matt may be overstating the degree to which a race neutral class politics will dominate at the polls- for many of the excellent reasons you point out. I don't think he's overstating the degree to which those politics are better for equitable outcomes RELATIVE to a race forward politics. Obamacare was certainly demonized and not fully implemented. It also passed and was able fend off repeal. Compare that with, say, Cory Booker's "Baby Bonds" plan which was a universal policy sold as way to combat racial inequality- that didn't even get traction in the Democratic primary. Now Booker was a weaker candidate and it's a little unfair to compare these two things. Still, I think it points to the fact that even if a more class oriented politics isn't a cheat code to super majorities, it seems like a clearly better move to sell your policies as benefitting the broadest possible audience.
I do think you raise many excellent points about how important it is to confront racism in America. However, I find myself in a bit of Catch-22 thinking about this: the more central you think racism is to American politics, the more backlash you should expect towards explicitly anti-racist policies. Shouldn't a firm belief in the centrality and power of racism lead a pragmatic politician to be even MORE circumspect about cloaking their racial equality policies under the cover of universalism? That is an answer that feels like moral cowardice, but how wrong is it? Discretion being the better part of valor and all that.
“ To be a bit clearer, what’s missing from this argument that makes it much weaker is that if you do this analysis against a whole range of socioeconomic outcomes around health, education, criminal justice, civic engagement, social capital, etc. there are huge racial gaps across class there as well. And living under those conditions is disempowering and miserable.”
But doesn't this argument miss class arguments like Florida (which Trump won) voting to increase minimum wage? To quote a very smart post, that's "what’s missing from this argument that makes it much weaker."
People vote differently when it comes to direct democracy and narrow referendums than when it comes down to politicians with their own identities who also push the same ideas. Democrats running on Medicaid Expansion still lose in many red states, Medicaid Expansion absent of any Democratic identity, somehow, wins.
The fact that the Deep South (along with some mountain states too!) did not agree to Medicaid expansion may or may not have to do with racialized politics. I suspect not; I doubt they rejected it because they didn't want their Black residents to benefit. I think it goes more to their whole philosophy of government, whether or not the beneficiaries would be white or some other race.
Medicaid Expansion (like minimum wage raising) easily gets tied up in the Identity of the politician(s) who push / support it...and absent those direct connection (on a ballot initiative, for instance) they drastically over preform politicians from the Democratic Party who run on the same things in the same state. In the Deep South (where i am from), there is one thing that’s synonymous for most folks when it comes to Democratic politics in these states (the identity of who wins elections in those Democratic districts)...and while that might not be explicitly whats putting people off...it’s rather hard to ignore. It’s also complicated by rich / white people often use dog whistling about racial redistribution because it’s the quickest shortcut to undercut anyone trying to get around those identity issues (say like a white Democrat who isn’t really running without any messaging that even mentions race)... and so the true or aggregate reasons often get very hard to suss out. 😣
This post triggered some thoughts about how potentially malicious corporate wokeness can be. Not to get too tin foil, but an especially cynical corporate PR department could be promoting cultural wokeness to both win kudos from cultural elites while spiking the football on risking real change (especially economic change) because they know tying woke goals to liberal economics poisons both politically. Have their cake and eat it too!
Potentially? It seems plain as day to me. Indulging your football metaphor, look at the "END RACISM" painted in the end zone of an NFL game for a particularly silly example.
Corporations are amoral entities formed to enrich shareholders. And I'm not saying that's inherently a bad thing! But to expect anything more from them is folly.
I’m sorry, but some of this conflates two basic facts:
1. The US has massive wealth inequality with an increasing concentration of wealth by the super rich.
2. There has been a persistent racial gap in median household wealth for decades. As of 2016, the median net worth of white households was about ten (10) times that of the median black household. [1]
10x is enormous! This demonstrates that there is a much broader set of white households that have a moderate amount of wealth. The racial wealth gap is persistent across educational levels, across age groups, and across income quintiles! [1][2]
There is a long history - even just looking in the 20th century - of the US government providing wealth to white families and excluding Black families: the GI Bill, the New Deal largely exempting agricultural and domestic workers, promoting nationwide segregation through restrictive covenants and redlining for federally backed mortgages, etc.
Yes, I agree that suburban districts have continued to remain segregated largely through single-family zoning and land-use restrictions, and true integration would include more multi-family housing.
Thank you! Matt's political advice here is reasonable, but his statistical argument is not. If the claim is that outliers are skewing the comparison, then using medians is a lot more informative than excluding an entire quartile of one population.
I don't have strong statistical chops, but I read the argument to be, essentially, that there is a discrete capital-owning class that is nearly all non-black, and outside of that class changes in wealth over time don't differ between whites and blacks (which suggests that race alone isn't why existing wealth gaps aren't closing- a currently poor black person and a currently poor white person, on that theory, are equally unlikely to accumulate wealth)
What's the statistical problem with bucketing the capital-holding class, so defined, and comparing the remaining white and black people?
I don't have strong stats chops either, but I think the current analysis is question-begging. It provides evidence that there is strong over-representation of whites at the very high ends of the distribution, but it then chops off 25% of whites from the analysis, not just billionaires.
To make that valid, you'd need to find what is the 75th-percentile wealth for white individuals, and then also remove black individuals above that wealth value, or provide evidence that the number that would be removed is small enough that it wouldn't alter the analysis.
I'm not sure you would have to do that. A lot of people in these comments are hung up on the asymmetry of only removing the top white quartile, but I don't think that's especially relevant, because the point isn't to compare all whites to all blacks. It's to show that there are a significant number of white individuals whose experience matches those of black individuals.
Yes, as you move up the wealth ladder, you will see more and more white people, but that's separate from the 75% argument. The key idea there is that if you don't look too far up the ladder, what you'll see is a whole lot of white people in very much the same situation as their black counterparts.
Except the difference matters. Because if the median is lower for blacks, then programs targeted at the middle class--e.g. student loan debt forgiveness, or non-refundable tax credits--will help black folks less than white.
The way I see it, what we really want to be doing is helping as many people who need help as possible, starting with the ones who need it most. The fact of the matter is that focusing on race will cause you to systematically ignore huge swaths of people who need it, and the 75% graph shows that.
It only seems weird to only use 75% of white people because you are starting on the assumption that race is the most relevant way to compare groups. I'm not sure that's true, especially if our fundamental goal is to make society, and the individuals that inhabit it, better off.
No, I think that misses the point of the theory (or at least *a* theory).
To: wit, if anti-black animus/prejudice was major factor currently hindering black wealth accumulation, you'd expect to see it bite in the form of less wealth accumulation in black people vs. white people at every level of wealth.
If, conversely you found that
a) at $0k, $10k, $50k, etc. of net wealth, black and white households had exactly the same average changes in wealth over time, and
b) at some high wealth threshold where there were very few black people (say, $500k) changes in wealth over time looked very different from households lower in the spectrum
that would seem to me to suggest the current racial prejudice is not the variable driving current difference in wealth accumulation--class (being above that, say $500k threshold) is. (Although earlier racial animus would obviously to blame as to why they are so few black people at the wealth level that is decoupling).
Whether that's a correct theory seems like an empirical question I don't have a take on, but the logic makes sense to me. The argument isn't about making an apple-to-apple comparison of aggregate black wealth and white wealth.
> if anti-black animus/prejudice was major factor currently hindering black wealth accumulation, you'd expect to see it bite in the form of less wealth accumulation in black people vs. white people at every level of wealth.
Two problems there:
1. Matt's stats don't show that is not the case, because increases in wealth by black folks towards the top of the distribution could in fact be covering lower accumulating by blacks within the same wealth range as the bottom 75% of whites.
2. More importantly (IMHO), if *historical* prejudice is a major factor, as reflected in the current median, then increasing everyone below some threshold by the same percent or even amount would still leave black folks behind due to historical discrimination.
> 1. Matt's stats don't show that is not the case, because increases in wealth by black folks towards the top of the distribution could in fact be covering lower accumulating by blacks within the same wealth range as the bottom 75% of whites.
Sure, the chart he put up is consistent with/suggests the scenario I laid out might be happening, but you'd need a more fined-grained analysis looking at actual wealth levels & wealth changes to be sure.
> More importantly (IMHO), if *historical* prejudice is a major factor, as reflected in the current median, then increasing everyone below some threshold by the same percent or even amount would still leave black folks behind due to historical discrimination.
Yeah, this seems obviously true and correct, but is a bit orthogonal to the original discussion. The gap in wealth levels is inarguably the result of historical animus againsts/repression of black people. Increasing everyone in the bottom's 75%'s wealth by 10% would still leave black people within that 75% behind, purely as matter of math.
But if that increase was accomplished by radically lowering the wealth of the top 25%, the overall white-black gap would shrink because of the racial composition of the top 25% being so non-black.
I think your reasoning is sound. I think that Matt's proposals for policy are good, but he used very shaky statistical reasoning to support his thesis about the racial wealth gap.
The design of policy to narrow overall wealth inequality should absolutely be focused on helping those in lower quartiles of wealth distribution. The baby bond policy in NJ is a good example of this type of policy - https://www.nytimes.com/2020/08/25/nyregion/baby-bond-nj.html
I'm frustrated because Matt cited a pew research study that refutes his central thesis. Here is a direct quote: "Among lower- and middle-income households, white families have four times as much wealth as black families and three times as much as Hispanic families."
I encourage all to read it. It's got very straightforward charts and graphs. No strong stastical chops needed.
The claim, as I see it, is not necessarily that outliers are skewing the data, but rather that there are a significant number of people with analogous situations between the two groups.
There's a pretty big difference between "there are a significant number of people with analogous situations" and "the situations are essentially analogous except at extremes." I read the article as suggesting the latter; but I agree with you that the former might be enough to justify the political advice--which again, I think is reasonable.
The question of how to represent an average is a well-tread topic in statistics. Using a mean when you know the distribution of your set is heavily skewed by extreme outliers (super wealthy) is a misleading choice. At the very least you should compare mean and median to get a sense of the skew.
Matt Bruenig writes "if the bottom 90 percent of black families were given the exact same per-household wealth as the bottom 90 percent of white families…77.5 percent of the overall racial wealth gap still remains" and "97 percent of the racial wealth gap is driven by households above the median of each racial group." https://jacobinmag.com/2020/07/racial-wealth-gap-redistribution
This is very interesting, I'm very curious how this graph and Matt's graph can both be correct. I mean obviously they *can* be but it makes it seem like the precise distribution of black and white wealth deserve more scrutiny.
I think the proper way to illustrate this would be to take away the top 25% by wealth (regardless of race) and see what the racial disparity in wealth looks like in the bottom three quartiles. I assume it's much more pronounced than in the gap presented.
And yeah, focusing on the bottom three quartiles isn't indefensible, but, I mean, also someone at the 75th percentile of wealth is not particularly wealthy. If racial wealth disparities were entirely focused in the upper class, like if you got essentially no racial wealth disparity once you were looking at the bottom 95%, then sure, that makes the case that no amount of focus on racial wealth gap would meaningfully improve the lives of the overwhelming majority of black people. As-is, it does seem clear that a hypothetical program might reduce the wealth gap by specifically targeting black people in a way that affects the lives of tens of millions of everyday people.
None of that is to disagree with the fundamental point that just targeting programs that increased the wealth of the bottom 25%, or 50%, or 75%, would both a: reduce the racial wealth gap, b: be better politics, and c: help poor white people who do actually need help!
I think somewhere we got told "means are bad, medians are good," but medians can be misleading as well. Like if the deal is that 40% of white people have $0 wealth, and then it rapidly increases from there, and 50% of black people have $0 wealth, and then it rapidly increases from there, then median white wealth is plausibly $80k and median black wealth is $0 and it looks gigantic, but obscures that actually very similar percentages of both groups have $0 wealth. Not that this is necessarily actually what's happening! I don't know! But just contra "we can use medians for this."
So ever since I read about this paper funded by Nick Hanauer I have just been gobsmacked by how much concentration of wealth at the top has been harmful to everyone else and caused most of the resentment and anxiety fueling a ton of our current social and political ills: https://www.rand.org/pubs/working_papers/WRA516-1.html If productivity growth had been shared more equitably from the mid 1970s to now all boats would be higher (they find median income would be doubled!) -- and hopefully we could focus more on eradicating racial and gender and other sources of inequality and discrimination. And the rich would still be plenty rich -- just not insanely so. I am super pro capitalism because markets do a good job of improving our material standard of living and pushing innovation, but we need to redistribute our GDP more equitably or we will never shake off all the conspiratorial crazy -- because there actually is a very small elite that controls an unhealthy amount of wealth.
"If productivity growth had been shared more equitably from the mid 1970s to now...And the rich would still be plenty rich -- just not insanely so."
So you didn't really mean "if productivity growth had been shared more equitably," you meant if more money had been taken from the most productive people and given to the less productive people.
Given that a big part of the spike in inequality arises from a declining income share for labor and an increasing share for capital, ridiculous to gloss reforms as "
"more money had taken from the most productive people and given to the less productive people".
The issue is that even as workers get more and more productive, the additional income from that output is going to owners (many of whom are passive, non-managerial owners).
“The issue is that even as workers get more and more productive, the additional income from that output is going to owners (many of whom are passive, non-managerial owners).”
If capital becomes more important why shouldn’t a higher share of returns go to investors? (And investing is never passive.)
I don't see how the concept of "capital becoming more important" makes any sense, analytically. All productive activity requires both capital and labor; they're both equally necessary and "important".
The relevant question is "how does the productive surplus get split between capital and labor", and that's seems like a question of economy-wide and society-wide structural factors, not some nebulous concept of "importance".
"All productive activity requires both capital and labor; they're both equally necessary and 'important."
Oh, you couldn't be more wrong. Imagine a line worker, in a factory, producing 20 $1 widgets an hour using a Widgetron-1000 that is owned by the firm that owns the factory. Further imagine that the worker's compensation (wages plus benefits) is $15 per hour. The firm's operating margin* is $5 an hour, so a 75/25 split labor to capital.
And then one day the firm buys a new-and-improved Widgetron-3000 that allows the same worker to produce 100 $1 widgets an hour.** The total output of the worker+machine (labor+capital) goes from $20 per hour to $100 per hour. Same input from the worker (1 hour) but five times the output for the worker+machine. This is clearly a situation where the labor and capital are not "equally necessary and 'important." Why, in your view, should we expect to see the same 75/25 split? (And before you answer, consider: If the returns to capital stayed the same why would the firm make the investment in the Widgetron-3000 in the first place?)
*This is obviously leaving out a lot of other production factors, from electricity to regulatory burden to maintenance to payroll taxes to opportunity costs. But, you know...widgets.
**Another oversimplification, of course, because the firm might instead decide to start selling the widgets for $0.50 each in order to increase market share.
Regardless of how productive the machine+worker combo is, the machine is worthless without the worker to operate it. The productivity of the two together defines the size of the surplus to be split between to two factors, but says nothing about what the division will be or should be.
The actual division of the surplus is a function of the leverage the two sides have vis each other, not marginal productivity--if, say, the factory unionizes shortly after the new machine is installed, the split that goes to the worker may well increase notwithstanding the vastly increased output facilitated by the machine; just as a prolonged weak labor market might drive down the worker's share even without any change in machines.
And although you're right that the firm wouldn't buy the new machine unless they thought they could extract enough of the surplus to cover the cost of the new equipment, that correct description of the firm's incentives doesn't describe how the surplus will or should be split once the new machine is in place.
I take it, then, that your response to my question, "If capital becomes more important why shouldn’t a higher share of returns go to investors?" is, "Of course it should
Well I mean that if the distribution of GDP across the income spectrum looked more like it did from 1945 to 1974 then we would have much higher wages for all but the top. I don't think that productive people were being shortchanged in that period of our history.
“I don't think that productive people were being shortchanged in that period of our history.”
Neither do I. Nor are they being shortchanged today. The reasons why “the distribution of GDP” (presumably you mean income) differs today are mostly pretty well understood: increasing returns to education, assortative mating and the resulting marriage patterns, etc.
"I don't use anything I learned in school at my job."
I highly recommend the book 'The Case Against Education' by economist Bryan Kaplan. Among other things he points out is that not only do most employees not use what they learned in school, but that most students don't even *retain* what they had learned. And he backs that up with data.
One of Kaplan's central points is that most valuable aspect of a college degree is not the education received but the credential. And that's pretty much what you seem to be saying about the "door in an elite world."
I don't know why you believe there wasn't money for elites 30 years ago.
“It's tautological to say that: (1) productivity is defined by market outcomes, and (2) market outcomes are fair because they reflect productivity.”
Note that I did not say such a thing. Though it is true that market outcomes generally reflect productivity. (And, no, productivity is not defined by market outcomes. Productivity is defined as “the amount of output produced by a unit of input.” [Baumol/Blinder, 4th Edition])
“For example, non-competes are banned in California, which makes it easier for workers to switch jobs and tends to make market outcomes more equitable.”
(2) is also pretty contestable! Given what we know about diminishing marginal utility of income, there's a strong case that we'd be better off doing more redistribution than we currently do.
Many (including me) agree that most people aren't rich. But what you don't address is that the history of universalistic programs is rivened by white racism. The history of social security is well known, But even the most recent attempt to benefit everyone--the Affordable Care Act--was opposed by white people everywhere, especially at first. It required a supermajority in the Senate and a Black president to be enacted--and even so Tea Partyists used its passage to elect an ACA opponent in MA. In short, what do you propose to do about white people who oppose universal programs mainly because they think universal programs benefit non-white people? Asking for me.
"In short, what do you propose to do about white people who oppose universal programs mainly because they think universal programs benefit non-white people?"
They are exceedingly few in number, so I suggest ignoring them.
As long as some percentage of opponents can be turned into supporters by focusing on class, we outvote the rest and wait for their (likely less racist) children and grandchildren to start voting.
As someone mentioned in yesterday's discussion, election victories and the policies they produce are built on small margins.
So I’m going to preface this with the acknowledgment that in the 90s Clintoncare similarly failed to achieve support for a universal program despite not being pushed by a white administration (granted with Clintoncare you had to dual problem of Hilary and sexism as well as the very very racialized crime reform driven by the very real crime problems of the late 80s/early 90s).
(And further caveat: I wasn’t really alive for all this, so my knowledge is all filtered through reading about people reflecting on this more than 20 years later)
So all that being said, I’m actually very curious about the counter factual scenario of a boring milquetoast centrist white democrat proposing Obamacare in the similar landscape of post 08 crash and huge democratic majorities? What happens then?
Could the Republican leadership (even if they were still determined to oppose) have held their caucus?
Maybe, but I doubt it because the base motivated some of the solidarity.
Would that crazy base that said they hated socialism even if it was pretty obvious they were motivated by racial grievance have come out? (For proof: see how they loved Sarah Palin’s true American schtick in fall 08 and the seeds of birtherism already coming out that fall)
Of course, say I am right, what to do now? Because this posits a very dark picture of a significant portion of the American electorate which doesn’t portend great things for us as a multi-ethnic democracy. How do we solve the problem of fear of the other, which is almost exclusively coded as black or brown? And Idk. It’s a pickle.
"So I’m going to preface this with the acknowledgment that in the 90s Clintoncare similarly failed to achieve support for a universal program despite not being pushed by a white administration (granted with Clintoncare you had to dual problem of Hilary and sexism as well as the very very racialized crime reform driven by the very real crime problems of the late 80s/early 90s)."
Just as opposition to Obamacare wasn't about race, opposition to the Clintons' attempt at health care "reform" wasn't about sexism. Rather, opposition was primarily about resistance to handing over so much power to a massive new bureaucracy and requiring employers to provide health insurance for all employees. (Let me know if you are unclear about why both of those were - and are - terrible ideas and I'll fill you in.)
Of course it was about race, at least in part lol. Obama being depicted as a witch doctor wasn't some high minded statement about government bureaucracy. It's a racial anxiety trigger.
I don't have strong views on Clintoncare or whatever, I prefaced with it only to head off it the argument that Republican opposition would have been as united in 2009 as the 90s even if Obama hadn't been there to stoke the racial anxieties of white americans. My argument is there is a good chance that it would have been different. Because 2009 was not the early 90s. The political realities were different and without the racialized opposition coming from the base, the republican caucus might not have held
This is being done by 40% (or less) of the country that Democrats will never get, no matter what they do. We have a bad habit of focusing on them, as if they represent everyone who voted Republican. The 8% who got Trump.from 40-48%? These are the voters who matter. And they're not birthers.
"The history of social security is well known, But even the most recent attempt to benefit everyone--the Affordable Care Act--was opposed by white people everywhere, especially at first."
I can only guess at this point, but it looks to me like there were a lot of issues in play during that election, including some missteps by Coakley. As Barack Obama said in an interview with New York Magazine:
"Well, the first thing that’s happening is I’m talking to Rahm [Emanuel, then chief of staff], and Jim Messina [then deputy chief of staff] and saying, “Okay, explain to me how this happened.” It was at that point that I learned that our candidate, Martha Coakley, had asked, rhetorically, “What should I do, stand in front of Fenway and shake hands with voters?,” and we figured that wasn’t a good bellwether of how things might go."
It seems to me that programs like public education, Social Security, Medicare and Medicaid have been extremely beneficial for black and white people. I'm not sure if I'm going to do anything to people that disagree with me regarding programs, however, but continue to advocate for ideas that make a ton of sense.
Unless I've totally misread the take here, what this (very good and informative, by the way) post represents is the intersections of Matt Yglesias and Bernie Sanders.
I think taking off the top 25% is statistically silly. If a baseball batter is one for four on average, he's mediocre. Make him one for three, and he's one of the best hitters in the game.
If you don't like that part, then just look at the part about how the top 10% is whiter than the general population, the top 1% is whiter than that, and the Forbes 400 list is even whiter.
The more elite the economic group you look at, the whiter it is. That's a true fact and it impacts the statistical aggregates. But diversifying the billionaire class would do nothing to help the typical Black family.
I agree that somehow "manufacturing" more black elites is performative, not substantive. To me, mobility is the key..
Raj Chetty has shown the racial disparity in mobility across generations. To your point about land use, Chetty has also shown a very positive correlation between mobility and time spent growing up in a more socio-economic diverse zip code. That's more important than almost any other factor. So affordable housing in your MA town example would be exactly on point
TL; DR Please use histograms when discussing distributional issues.
David Roberts has a point that lopping off the top 25 percent of the distribution is silly. That said, Kevin Carney (and by extension you) have good reason for wanting to do it; to only show the means (medians) does not tell you a whole lot about how two distributions compare because you are only reporting one moment (quantile) of each distribution. However, the better solution would be to show ALL of BOTH distributions (i.e., by using a histogram or nonparametric estimate of the density), like Pew Social Trends does with income here: https://www.pewsocialtrends.org/2018/07/12/appendix-a-income-distributions-of-whites-blacks-hispanics-and-asians-in-the-u-s-1970-and-2016/.
The only downside is that you need more figures (and therefore more space) to show how the comparison changes over time. In this case, however, it seems like the emphasis of the story is the comparison of the two distributions, not the time trend of that comparison, so this shouldn't be a big deal.
It also wouldn't help the typical white family as well. This is one of the things that conservatives do so well when they show their black politicians or supporters and make white voters feel good without actually doing anything for them. It creates this veneer of diversity while at the same time letting the politicians play on racial resentment to win political races. If poor white voters were instead united with poor black voters in fighting for things that actually improved their lives you could have a very powerful political party. Poor white people feel that they are connected to rich white people in some weird convoluted way even though they are constantly getting taken advantage of (i.e. Trump's election defense fund that was just used to pay off campaign debts). Doing the same thing for black people by making them somehow connected with black elites plays on that same racial cord that Leftists criticize conservatives using and does nothing to actually solve underlying problems (it actually makes them worse i.e. Obama winning the White House and the backlash that ensued)
Sure but poor white people are still very closely connected with race. It's not like they are uniting with poor black people to form a big coalition against white, wealthy people. They are still voting for conservative policies that keep white, wealthy people in power (i.e. Donald Trump) even though it goes against what would actually help them (such as Medicare and welfare). So I see your point that they may not have a connection with the new wealthy white people but they still believe they are better than poor black people.
This is the C Vann Woodward thesis and the problem is that he wrote that in 1955. I question whether there is not something else going on in the more than half a century since then. What I see is an egalitarian culture gone rancid. Trump people hate liberals, not blacks.
I tend to agree cash payments are unworkable. But I'm sympathetic to more targeted measures that address more narrow systemic examples. So here in Chicago I would support a program that increased or even re-distributed school funding to previously redlined neighborhoods in proportional amount to cumulative loss in tax revenue from limiting the capital appreciation of the housing assets. But maybe Chicago was so egregious here in their application of redlining that I'm biased vs. the national view.
I think Matt would argue that a simple formula change to allow for more funding in poorer districts (which is in many cases what we have actually done) would be easier to justify to voters and accomplish the same thing.
That sounds like an incomplete analysis. Giving poor people money lets them spend it, so it would improve the life of anyone who can sell things to them.
I didn't interpret it as anything statistically rigorous, I think it just illustrated that the economic experience of the vast majority of white Americans is very similar to the economic experience of Black Americans. I find it a useful framing to keep in mind.
I think that it’s worse than that. There is an abundance of empirical data that demonstrates persistent racial wealth gaps measures by median wealth. The statistical trick employed uses mean wealth, so the ultra wealthy black population’s wealth is included while that of the white population is not. It doesn’t compare apples to apples.
Comparing median wealth would be more representative of “typical experience”, which they did not do.
The pew research that Matt linked shows persistent racial wealth gaps across income and education levels!
The homeownership rates are persistently higher for white households, which supports the story of a broad based wealth gap and agrees with the documented history of government policies that made it easy for white families to get a federally subsidized mortgage while preventing black families from doing the same for decades.
It's still relevant though because at the end of the day we are still talking about individuals. Your numbers are important, too, but the fact that you can find huge numbers of white people in the same situation as analogous black people does imply that race may not be the best thing to focus on in order to enact meaningful change.
I'm confused about the mean/median difference. Given the right-skewed distribution of wealth in the US, I would expect the gap in Black/white median family wealth to be smaller than the gap between their wealth at the mean. But according to a Fed publication, that's not true: Black family wealth is 15% of white family wealth at both the mean and the median.
My point about focusing on medians is that this is a simple way to negate the skewing effect of outliers, which is presumably what the impetus behind lopping off the top 25% of white wealth was. It's a weird thing to do.
The point of looking at medians is to compare the 50th percentile of white families to the 50th percentile of Black families. The fact the the gap persists is precisely the point that I was making - the racial wealth gap is not just the result of there being more super-rich white billionaires.
To your specific point: the link you provided shows a slightly larger gap for when using medians - Black median wealth is 12.8% of white median wealth, while Black mean wealth is 14.5% of white mean wealth. What I take from this is that wealth is right-skewed both among Black families and white families.
I agree with you! My confusion was that I expected there to be a much bigger bias when using means instead of medians. And that turns out not to be the case.
Just want to echo that you're right, RS. Use of median in skewed distributions is 101 stuff. Love Matt, find the political tactics here reasonable to consider, but this post's premise is innumerate at the core
That's not analogous, though. A baseball batter is a single person, and the average is a single number intended to summarize his ability. Looking at individual hits by themself would actually be deceptive, since you want to know their performance long term.
The case of a group's wealth is more complex than that. "White people" are a group of individuals, and what we really care about are the individuals, not the aggregate. For a white person in the bottom 25%, the wealth of the top 25% does them no good, and a statistic like the one Matt presented helps illustrate that point.
Another way to think about it is this: dividing by race is inherently arbitrary, you could just as easily divide by hair color, or favorite pizza toppings, and the distributions would still be valid; "the bottom 75% percent of white people" is potentially just as valid a group as any other, as long as you contextualize why you're using that metric. On the other hand, there's no alternative way to divide a baseball player's hitting average.
CSB: I used to be a union organizer. One of the (many) things that we had to fight was "bosses"/conservatives driving a racial wedge between people who were fighting the same fight. And now " liberals" do it; if you even attempt to explain why this is a bad idea, they will call you racist.
Like many things that can seem incidental or coincidental, this isn't. The kind of people who fund the CAP (for instance) would much rather hire three black women and have everyone tell them how awesome they are, than give 300 people a raise (including *30* and genuinely make a
All of the "just lop off the richest quarter of white people" analysis seems very suspect to me. Excluding rich whites and not excluding rich Blacks would be expected to make a wealth gap disappear, right?
What happens if you re-run all this analysis after also lopping off the richest quarter of Blacks. I suspect that you'll find a racial wealth gap again.
I'm not going to look this next bit up to verify it (poor form on my part, but it's very early in the day), but I remember reading once that the median net worth of Black families in Boston was effectively zero. The median wealth of white Boston families is presumably not zero.
Idk man. I think sometimes people try and reduce everything to class because "bring in the tide to lift all boats" is an easier sell than "bring in the tide but also plug the holes that people sawed into the hulls of 12% of all boats".
I'm always glad to read what you've written though, especially when we disagree.
Page 21 of the paper Matt linked has a graph of the wealth gap between the median white household and the median black household. It shows median white household wealth bouncing between $60k and $100k 1989-2013, with an upward trajectory for the 90s and 00s offset by a big fall with the Great Recession, ending up at $80k. Meanwhile, median black household wealth started the period at $10k, rose to nearly $20k in 1998, but declined every year after to almost $0 in 2013.
I think you're on to something but the solution would not quite be the one you suggest.
You could look at the wealth level of the 25% percentile (for example) of whites, and then "lop off" anyone (regardless of race) above that level from the analysis. This would be the proper way to investigate this question with the, um, "lop off" mechanic.
If Matt is correct, this would almost entirely remove rich white people from the data and his point would hold. In theory though, compared to the analysis he describes, it could instead remove enough rich black people from the data that the racial wealth gap would be evident.
(I believe the other evidence for his point is strong, to be clear, but I do agree that something about this particular point of analysis seemed a little off.)
I don't disagree with what you've said, but at the end of the day if all you do is focus on trying to lift all boats then (pending someone doing the analysis) you aren't going to get rid of (all) the wealth gap.
If that isn't your highest priority then no problem (it isn't mine either!), but that's different than chalking up the gap to just being a class gap in disguise.
I think some would argue that you are sort of side stepping the institutional racism argument. That argument would suggest there is an extra barrier on top of class that will still make it difficult for some people to rise economically even if you try to adjust some of the prosperity metrics through technocratic innovation. Do you have an opinion on how ingrained racism could be overcome directly rather than circumvented?
I mean as I said at the end, I think you need to tackle racial discrimination as an autonomous sphere of activity. If academics can do blind resume tests to detect labor market discrimination, then the Civil Rights Division of the Justice Department can use the same tactics to do stings.
We have some workable models of police reform. We can continue the process of promoting more diverse representation in political office.
In other words, I think most more or less "normal" ideas for fighting discrimination and racism make sense. But attempting to directly target the racial economic distribution is a confusing idea, given the underlying circumstances of severe inequality.
I see where you're coming from, but this sounds a bit too authoritarian for my taste.
Also, depending on organizational goals and culture, being able to identify gender and racial cues on a resume may actually increase hiring of minorities.
This is anecdotal, but I ran an engineering organization and we were always behind our diversity goals. I'm a bit embarrassed to admit it, but if I'm being honest, whenever I saw a resume with a women's name on it, my filter changed from "is this person an outstanding top of their class engineer" to "is there a way we could possibly make this person work in our organization."
"When resumes were made anonymous, participating firms became less likely to interview and hire minority candidates. The gap in interview rates between non-minority and minority candidates widened by 10.7 percentage points, from 2.4 percentage points in the standard procedure to 13 percentage points in the anonymized procedure. At the hiring stage, the gap widened by 3.7 percentage points."
Interestingly, because the study was opt-in on the part of employers, the researchers' conclusion was that the companies eager to adopt a new diversity program tended to be the ones whose biases ran in the opposite direction to begin with:
"Why did anonymized resumes lead to fewer interview offers to minority job candidates? The answer may lie in the difference between firms that selected to participate in the program and those that did not. Participating firms that were randomly assigned to the comparison group (and thus continued to receive resumes with all identifying information) were twice as likely to offer interviews to minority candidates. The average firm volunteering to participate in the study thus appeared to positively value minority candidates’ resumes in general, and anonymization changed the decision-making process for these firms."
My take is that imposing a legal bureaucracy in charge of standardizing resume inputs is not going to get us what we want. But we can and should definitely advocate that companies or hiring managers do that.
To offer a personal story, I recently hired from a pool of about 50 applicants and had an excel file to pick who I wanted to interview from. I hid the applicant names and university names and only evaluated each candidate based on their credentials like degrees, certificates, and listed technical skills or knowledge.
From this "blind" process, I picked 7 people to interview: 1 black female, 1 Asian female, and 5 white men. I ended up hiring of the white men who was clearly the best candidate, but I was pleased that my little process seemed to have generated an equal-opportunity process.
My understanding of the classic blind-auditions study is that actually the original data analysis was bad and it was a null result, and no one has ever produced a successful replica study.
Female orchestra players increased significantly after the policy was implemented. Occam's Razor suggests that the policy was the reason for the increase. I think the argument that audition panels suddenly had a change of heart is quite a bit less plausible.
One way to test this presumption would be to conduct a statistical analysis of the effects of the policy change on a per-audition basis, adjusting for changes in the relevant populations. Turns out that, when you do this, you don't find any substantial evidence that the policy improved the outcomes of female musicians, and some evidence that it may have hurt their chances.
The advantage of doing this, rather than reaching for Occam's Razor, is that it helps to disentangle the effects of a single policy from the more general societal changes which have boosted female employment outcomes in most industries over the last fifty years without the adoption of blind auditions.
I think instead we just need to make those traits not as valuable. We need to change the biases people hold rather than forcing bias to not be in the process. How would you prevent gender from being an issue in an interview? Would you just black out the person's face and distort their voice? We need to change the stereotypes we hold to be more productive in society rather than trying to get rid of stereotypes altogether. I'm not sure exactly what types of stereotypes would be most beneficial but I do know that trying to remove a crucial part of the human brain is a very difficult and arguably impossible road to go down.
I pitched a seed round start up for this. The idea was a CarFax like solution for HR departments. It would autonomize resumes prior to submittal. I still think it's a good idea.
White guilt isn't completely worthless but you are right that it isn't the best way to get things done. I think this is one of those "two things can be true" issues where we can probably find sensible ways to address both issues but it would take the kind of consistency of message and coordination that the current Democratic Party lacks. Maybe Matt is purposefully circumventing that part of the argument because he realizes it is divisive and he is trying to cut through to a solution that could help in a very utilitarian way.
Bo can you share the value of "white guilt?" I've seen a lot of evidence and discussion about how attempting to shame is counterproductive. Would be curious how you see this being an exception? Especially if it inclines poor whites to object to economic redistribution?
In the "value" order I certainly don't put it above lots of other useful motivations. I do think a small amount of guilt or shame for things in our past can be useful as long as it is properly contextualized and the resulting emotion is put toward accomplishing good in the world. I do not think it should hoisted on a brand banner for the whites on the left to rally around, I think it has a small personal motivating use for some people. That's really the extent of it. So not useless but also not something that is useful to champion.
But the "makes them feel good" is the inherent moral imperative that comes along with a lot of stuff we do in our communities. We like to do stuff that we feel has real social benefit, that makes us feel good. I agree that the hazard is when this becomes a form of selfishness. Some people may be frustrated because they feel like lots of things have been tried already and they are not seeing results. We just need to be mindful of that headspace and respectful of the difference in peoples intentions.
It may be a commonplace now, but I’ll say that I live in a mostly white, professional neighborhood where you can see lots of BLM signs - and also lots of people who are ferocious opponents of more-dense housing or the housing voucher policies that would open the neighborhood to less affluent residents.
Good piece. I think you can make the same point about university admissions. I don't get why there are so many battles over racial diversity in college admissions. To me it makes more sense for colleges to simply allot more spots to lower income families and that should increase diversity along with it. Unless they don't really care about assisting lower income kids and would be happier taking in upper class minorities.
I think this ties into another, related question, one which maybe Matt would like to ponder in another post: why has the US -- pretty much alone among Western nations -- never had a strong, let alone successful, socialist movement?
Many have attributed it to an ingrained suspicion on strong governments. Probably true, but also true is that most Western nations have historically been fairly uniform ethnically, racially and mostly religiously; their major dividing lines have been on class lines. In the US, class never has been the major activating feature of group conflict, whereas racial and ethnic divisions have been. Thus the default tendency in the US -- both on the left and the right -- is to see things divided on racial and not class lines.
If that's deeply embedded in our culture, I suspect class-based policy initiatives may have trouble succeeding.
I think people are massively overstating the importance of naming conventions. The British Labour Party is technically a socialist party but Tony Blair and Gordon Brown saw Bill Clinton as a kindred spirit. Indeed, long before him, a Labour Leader said that "English socialism owed more to Methodism than to Marxism". All Europe's key socialist parties ultimately renounced Marxism, most decades before the end of the Cold War. America has a history of powerful trade unions, it has a history of public agitation for key utilities to provide through government funding, it has a left wing party that is closely aligned with trade unions, an alliance that has endured as the centre of gravity amongst those unions moved from heavy industry to the public sector.
Ultimately these arguments come down to "why did the American left not adopt the naming conventions inspired by arguments of a German economist who lived in London", I dunno, maybe because America is half the world away from either country. There's no American Tory Party either, but no one goes round saying that America is unusually left-wing.
Interesting but not really related to what I was saying. And that was that the primary cleavages in American politics have been on racial and ethnic lines and not class lines. I don't think you can say that about the Labour and Tory parties in England historically.
Actually there's a lot of demographic divides between the Tories and Labour/Liberals (basically English vs Celts, Anglicans vs Other Protestants, South vs North, Rural then Rural/Suburban vs Urban), and even between Labour/Liberals (Catholics vs Non-Anglican Protestants). Likewise Britain's BAME community has been a solid Labour voting block, not as much as African-Americans, but comparable to (say) Hispanic Americans. The actual difference is that the brief periods progressives get control in America, they can't push through as much radical change as British ones can, because our system has virtually no veto points. But that's also why our Thatcherites were able to privatise the buses, trains, air traffice control, post office, etc whereas Reganites couldn't.
I think Bernie was the most successful socialist movement, but I think it is the type of socialism that became popular in Western and (especially) Northern Europe: social democracy. I think its movement is stuck at the moment due to running poor campaigns, but I would be happy if a social democracy movement were to start to gain steam
I wonder if our increased income inequality (concentration of wealth with fewer people) will increase class dividing lines.
I'm pretty conservative, but I dislike wealthy people. If was king, I would tax Gates, Buffet, Zuckerberg, etc... so much they would be driving uber to pay bills.
Or at the very least, a 90% death tax on all estates values above 50 million.
I'm pretty pro founder billionaires - or certainly more so than your typical progressive. I think A LOT of the founder billionaires are simply the extreme end of the distribution. It's a biased sample. You're not measuring the millions of failed founders and I think there's a reason Silicon Valley exists here and not elsewhere and favorable tax structure plays some part.
WHAT I CAN'T AT ALL FATHOM is how much offshore tax dodging we allow and why we can't attack inter-generation wealth transfer. To me it's all criminal. It's hurts to see all the Walton heirs atop the Forbes list. I think Sam built an amazing business. I don't think his kids did anything other than win a lottery.
Agree. There's no greater equal opportunity cause than dramatically hiking inheritance taxes. Let founders who genuinely created something enjoy their wealth while they're alive, but when they die distribute it equally to everyone, not just a few undeserving people who won a parental lottery, and then think got it because they deserve it.
I am pro founder billionaires as well... ok... Im technically pro take a lot of their money while they are alive via wealth tax and take the rest when they die via death tax. And Im not even liberal. I just dislike rich people.
The US has had some very strong *socialists* (maybe not TR!) but not a strong socialist *movement.* And, yes, I define that as winning elections, implementing socialist ideology or maybe leading popular movements to take over government through other means.
Indeed, FDR incorporated thinking of some socialists. So did Bismarck under Staatssozialismus. Neither, obviously, was a socialist.
My politically conservative father has not voted for a Democrat since Jimmy Carter. This is a message that appeals to him. CEO pay & decreasing living standards in his area (very white, very rural): two things he talks about, constantly.
If the message is racialized, it becomes less salient. To his credit, he is fair-minded enough to acknowledge our history of racial inequity and how it factors into economic outcomes. Still, these conversations are meaningless for his community unless they are tied to economics, because they only have poor people that are white. Southwest Virginia not likely to go progressive any time soon, but I think this should be the mantle of the oft-malpracticed Democratic strategy to win in rural communities.
Completely agree. A widely popular Democrat would be a quite economically-focused one.
Some of your points relate to the narrow group of billionaires, while the top quarter probably includes most of your white audience. The slippage in who is "rich" is a big problem in framing solutions about class. I have a feeling the necessary solutions involve redistribution from the top quarter, but in the comments you are telling people they can ignore that point if they don't like it!
This question is raised (and purportedly answered) all the time but I haven't seen a lot of data to back it up, even though it's really important! I'd be very curious to see clear quantitative assessments of how much money you'd need to raise to meaningfully raise living standards of the bottom 50%, and whether you'd really need to tap the richest quarter for more taxes or just the top 10% or 1%!
Of course even as I write this I realize attempting to assess this in the objective way I describe is impossible in a world where interest rates are zero or even negative, so the proper compromise is to simply spend the money without raising taxes on anyone... The proposition of raising taxes for the *purpose* of addressing inequality *by taking money from rich people* is not an objective quantitative question.
The way the Nordics do taxes, the top tax bracket includes much of the upper middle class. It often starts at around 1.5x the average income. But that's also because they have fewer rich people to tax. https://taxfoundation.org/bernie-sanders-scandinavian-countries-taxes/
I agree with the political argument that it’s better for progressive coalitions to focus on economic redistribution and maybe public policy more broadly in a race neutral way. I also agree that better enforcement of laws that currently exist can go a long way in ameliorating discrimination.
That said, I don’t think there’s enough acknowledgement of how much US political economy is racialized, and how challenging it is to execute your strategy.
For example, Obama tried to follow this logic pretty closely, and we were barely were able to get Obamacare passed and maintained (to this day). And the policy impact of Obamacare is less likely to reach Blacker states like Alabama vs. whiter states like Massachusetts.
And this isn’t to say that Obamacare wasn’t the best that we could do. It’s suggesting that those design choices reflect a fairly racialized political economy even when you’re trying to do the opposite, which has the intended effects.
So the politics that we’re seeing is an effort to change the country’s political economy by not pretending that race isn’t a considerable animating factor for the behavior of political actors, but calling it out, and rejecting it as the only way to get out of our toxic political dynamics. It’s just not sustainable for the people living through it.
To be a bit clearer, what’s missing from this argument that makes it much weaker is that if you do this analysis against a whole range of socioeconomic outcomes around health, education, criminal justice, civic engagement, social capital, etc. there are huge racial gaps across class there as well. And living under those conditions is disempowering and miserable.
For example, economic mobility for middle class black boys is lower than for poor white boys! Rich black boys are more likely to end up in jail than poor white boys! That’s not a class problem. And the experience of living in that hierarchy isn’t properly reflected in the outcomes that we’re focused on.
An important question to consider, is how do you build a politics that doesn’t acknowledge that? Barack is right! Defund the police is a terrible slogan - but I am not sure “Reform the police” would have created better politics at the margins. It would be demagogued in the same way, and not forced a reckoning with the issues we need to have.
I am halfway convinced that your approach would work, but can we go to the places in the country that are closer to this approach vs. not and are their outcomes any better or any less racialized?
Most people of color don’t want to have every argument and analysis on the basis of race, and many are aware that wealthy POC can learn to live comfortably in an unsustainable class hierarchy. They also know the impact of race permeates everything and it has to be extricated in parallel with executing a more class based politics and policy.
Lots of excellent points, Nick. And to your point about racialized political economy, I'm surprised that Matt is not familiar with the arguments of many social scientists that a major reason for low support for a European-style welfare state in the USA is racial diversity: people are less likely to support redistribution if they view it as benefiting undeserving minorities. Trying to sell your programs in purely economic terms doesn't necessarily make the racial dynamics disappear.
An example of the argument that racial animus in the US reduces support for welfare-state programs, from Alesina Glaeser and Sacerdote: https://www.nber.org/papers/w8524
But the traditional Democratic approach - e.g. as recently as Obama, as Matt points out - has been to respond by downplaying race. The new Democratic approach of *highlighting* race seems likely to aggravate the dynamic described by Alesina et al.
Maybe within the college environment the direct approach to tackling racial animus seems far easier than in society at large?
A similar argument from a Canadian perspective: Joseph Heath suggests that it's human nature to divide into groups, but not necessarily by *race*. So in a multiracial society, you want to downplay the salience of race as much as possible. (For example, the coalitions represented by the major Canadian political parties are not divided by race.) https://induecourse.utoronto.ca/on-racism-and-race-consciousness/
Lol he's definitely familiar with it. He worked at a place called Vox. Because he didn't bring it up doesn't mean he's not familiar with it.
I think Matt may be overstating the degree to which a race neutral class politics will dominate at the polls- for many of the excellent reasons you point out. I don't think he's overstating the degree to which those politics are better for equitable outcomes RELATIVE to a race forward politics. Obamacare was certainly demonized and not fully implemented. It also passed and was able fend off repeal. Compare that with, say, Cory Booker's "Baby Bonds" plan which was a universal policy sold as way to combat racial inequality- that didn't even get traction in the Democratic primary. Now Booker was a weaker candidate and it's a little unfair to compare these two things. Still, I think it points to the fact that even if a more class oriented politics isn't a cheat code to super majorities, it seems like a clearly better move to sell your policies as benefitting the broadest possible audience.
I do think you raise many excellent points about how important it is to confront racism in America. However, I find myself in a bit of Catch-22 thinking about this: the more central you think racism is to American politics, the more backlash you should expect towards explicitly anti-racist policies. Shouldn't a firm belief in the centrality and power of racism lead a pragmatic politician to be even MORE circumspect about cloaking their racial equality policies under the cover of universalism? That is an answer that feels like moral cowardice, but how wrong is it? Discretion being the better part of valor and all that.
“ To be a bit clearer, what’s missing from this argument that makes it much weaker is that if you do this analysis against a whole range of socioeconomic outcomes around health, education, criminal justice, civic engagement, social capital, etc. there are huge racial gaps across class there as well. And living under those conditions is disempowering and miserable.”
So well put, thank you!
But doesn't this argument miss class arguments like Florida (which Trump won) voting to increase minimum wage? To quote a very smart post, that's "what’s missing from this argument that makes it much weaker."
People vote differently when it comes to direct democracy and narrow referendums than when it comes down to politicians with their own identities who also push the same ideas. Democrats running on Medicaid Expansion still lose in many red states, Medicaid Expansion absent of any Democratic identity, somehow, wins.
The fact that the Deep South (along with some mountain states too!) did not agree to Medicaid expansion may or may not have to do with racialized politics. I suspect not; I doubt they rejected it because they didn't want their Black residents to benefit. I think it goes more to their whole philosophy of government, whether or not the beneficiaries would be white or some other race.
Medicaid Expansion (like minimum wage raising) easily gets tied up in the Identity of the politician(s) who push / support it...and absent those direct connection (on a ballot initiative, for instance) they drastically over preform politicians from the Democratic Party who run on the same things in the same state. In the Deep South (where i am from), there is one thing that’s synonymous for most folks when it comes to Democratic politics in these states (the identity of who wins elections in those Democratic districts)...and while that might not be explicitly whats putting people off...it’s rather hard to ignore. It’s also complicated by rich / white people often use dog whistling about racial redistribution because it’s the quickest shortcut to undercut anyone trying to get around those identity issues (say like a white Democrat who isn’t really running without any messaging that even mentions race)... and so the true or aggregate reasons often get very hard to suss out. 😣
This post triggered some thoughts about how potentially malicious corporate wokeness can be. Not to get too tin foil, but an especially cynical corporate PR department could be promoting cultural wokeness to both win kudos from cultural elites while spiking the football on risking real change (especially economic change) because they know tying woke goals to liberal economics poisons both politically. Have their cake and eat it too!
Talk is cheap, and performative wokeness talk has no history of effecting costly change, so it's dirt cheap.
Potentially? It seems plain as day to me. Indulging your football metaphor, look at the "END RACISM" painted in the end zone of an NFL game for a particularly silly example.
The NBA going all out for Black Lives Matter, but will endanger their jobs by supporting Hong Kong. One of these things is an economic threat.
Corporations are amoral entities formed to enrich shareholders. And I'm not saying that's inherently a bad thing! But to expect anything more from them is folly.
I’m sorry, but some of this conflates two basic facts:
1. The US has massive wealth inequality with an increasing concentration of wealth by the super rich.
2. There has been a persistent racial gap in median household wealth for decades. As of 2016, the median net worth of white households was about ten (10) times that of the median black household. [1]
10x is enormous! This demonstrates that there is a much broader set of white households that have a moderate amount of wealth. The racial wealth gap is persistent across educational levels, across age groups, and across income quintiles! [1][2]
There is a long history - even just looking in the 20th century - of the US government providing wealth to white families and excluding Black families: the GI Bill, the New Deal largely exempting agricultural and domestic workers, promoting nationwide segregation through restrictive covenants and redlining for federally backed mortgages, etc.
Yes, I agree that suburban districts have continued to remain segregated largely through single-family zoning and land-use restrictions, and true integration would include more multi-family housing.
[1] https://www.brookings.edu/blog/up-front/2020/02/27/examining-the-black-white-wealth-gap/amp/
[2] https://www.fastcompany.com/90461708/why-wealth-equality-remains-out-of-reach-for-black-americans
Thank you! Matt's political advice here is reasonable, but his statistical argument is not. If the claim is that outliers are skewing the comparison, then using medians is a lot more informative than excluding an entire quartile of one population.
I don't have strong statistical chops, but I read the argument to be, essentially, that there is a discrete capital-owning class that is nearly all non-black, and outside of that class changes in wealth over time don't differ between whites and blacks (which suggests that race alone isn't why existing wealth gaps aren't closing- a currently poor black person and a currently poor white person, on that theory, are equally unlikely to accumulate wealth)
What's the statistical problem with bucketing the capital-holding class, so defined, and comparing the remaining white and black people?
I don't have strong stats chops either, but I think the current analysis is question-begging. It provides evidence that there is strong over-representation of whites at the very high ends of the distribution, but it then chops off 25% of whites from the analysis, not just billionaires.
To make that valid, you'd need to find what is the 75th-percentile wealth for white individuals, and then also remove black individuals above that wealth value, or provide evidence that the number that would be removed is small enough that it wouldn't alter the analysis.
I'm not sure you would have to do that. A lot of people in these comments are hung up on the asymmetry of only removing the top white quartile, but I don't think that's especially relevant, because the point isn't to compare all whites to all blacks. It's to show that there are a significant number of white individuals whose experience matches those of black individuals.
Yes, as you move up the wealth ladder, you will see more and more white people, but that's separate from the 75% argument. The key idea there is that if you don't look too far up the ladder, what you'll see is a whole lot of white people in very much the same situation as their black counterparts.
Except the difference matters. Because if the median is lower for blacks, then programs targeted at the middle class--e.g. student loan debt forgiveness, or non-refundable tax credits--will help black folks less than white.
The way I see it, what we really want to be doing is helping as many people who need help as possible, starting with the ones who need it most. The fact of the matter is that focusing on race will cause you to systematically ignore huge swaths of people who need it, and the 75% graph shows that.
It only seems weird to only use 75% of white people because you are starting on the assumption that race is the most relevant way to compare groups. I'm not sure that's true, especially if our fundamental goal is to make society, and the individuals that inhabit it, better off.
So I'll repeat what I've said a few times in comments now--I don't think Matt's political advice is wrong.
I think that in the interest of punching up his article, he made a stronger claim than the data warrants. That's all.
Fair enough. I can agree with that. I could feel as I was writing that I was drifting from the statements made by Matt anyway.
No, I think that misses the point of the theory (or at least *a* theory).
To: wit, if anti-black animus/prejudice was major factor currently hindering black wealth accumulation, you'd expect to see it bite in the form of less wealth accumulation in black people vs. white people at every level of wealth.
If, conversely you found that
a) at $0k, $10k, $50k, etc. of net wealth, black and white households had exactly the same average changes in wealth over time, and
b) at some high wealth threshold where there were very few black people (say, $500k) changes in wealth over time looked very different from households lower in the spectrum
that would seem to me to suggest the current racial prejudice is not the variable driving current difference in wealth accumulation--class (being above that, say $500k threshold) is. (Although earlier racial animus would obviously to blame as to why they are so few black people at the wealth level that is decoupling).
Whether that's a correct theory seems like an empirical question I don't have a take on, but the logic makes sense to me. The argument isn't about making an apple-to-apple comparison of aggregate black wealth and white wealth.
> if anti-black animus/prejudice was major factor currently hindering black wealth accumulation, you'd expect to see it bite in the form of less wealth accumulation in black people vs. white people at every level of wealth.
Two problems there:
1. Matt's stats don't show that is not the case, because increases in wealth by black folks towards the top of the distribution could in fact be covering lower accumulating by blacks within the same wealth range as the bottom 75% of whites.
2. More importantly (IMHO), if *historical* prejudice is a major factor, as reflected in the current median, then increasing everyone below some threshold by the same percent or even amount would still leave black folks behind due to historical discrimination.
> 1. Matt's stats don't show that is not the case, because increases in wealth by black folks towards the top of the distribution could in fact be covering lower accumulating by blacks within the same wealth range as the bottom 75% of whites.
Sure, the chart he put up is consistent with/suggests the scenario I laid out might be happening, but you'd need a more fined-grained analysis looking at actual wealth levels & wealth changes to be sure.
> More importantly (IMHO), if *historical* prejudice is a major factor, as reflected in the current median, then increasing everyone below some threshold by the same percent or even amount would still leave black folks behind due to historical discrimination.
Yeah, this seems obviously true and correct, but is a bit orthogonal to the original discussion. The gap in wealth levels is inarguably the result of historical animus againsts/repression of black people. Increasing everyone in the bottom's 75%'s wealth by 10% would still leave black people within that 75% behind, purely as matter of math.
But if that increase was accomplished by radically lowering the wealth of the top 25%, the overall white-black gap would shrink because of the racial composition of the top 25% being so non-black.
I think your reasoning is sound. I think that Matt's proposals for policy are good, but he used very shaky statistical reasoning to support his thesis about the racial wealth gap.
The design of policy to narrow overall wealth inequality should absolutely be focused on helping those in lower quartiles of wealth distribution. The baby bond policy in NJ is a good example of this type of policy - https://www.nytimes.com/2020/08/25/nyregion/baby-bond-nj.html
I'm frustrated because Matt cited a pew research study that refutes his central thesis. Here is a direct quote: "Among lower- and middle-income households, white families have four times as much wealth as black families and three times as much as Hispanic families."
I encourage all to read it. It's got very straightforward charts and graphs. No strong stastical chops needed.
https://www.pewresearch.org/fact-tank/2017/11/01/how-wealth-inequality-has-changed-in-the-u-s-since-the-great-recession-by-race-ethnicity-and-income/
The claim, as I see it, is not necessarily that outliers are skewing the data, but rather that there are a significant number of people with analogous situations between the two groups.
There's a pretty big difference between "there are a significant number of people with analogous situations" and "the situations are essentially analogous except at extremes." I read the article as suggesting the latter; but I agree with you that the former might be enough to justify the political advice--which again, I think is reasonable.
That is a great summary!
The question of how to represent an average is a well-tread topic in statistics. Using a mean when you know the distribution of your set is heavily skewed by extreme outliers (super wealthy) is a misleading choice. At the very least you should compare mean and median to get a sense of the skew.
Thanks for this. Wanted to put together something similar when I read the initial post this morning.
The original argument strikes me as bad math or bad faith for political expediency. Not sure which aggravates me more.
Matt Bruenig writes "if the bottom 90 percent of black families were given the exact same per-household wealth as the bottom 90 percent of white families…77.5 percent of the overall racial wealth gap still remains" and "97 percent of the racial wealth gap is driven by households above the median of each racial group." https://jacobinmag.com/2020/07/racial-wealth-gap-redistribution
Using mean (average) is problematic for obvious reasons, but rather than just ignoring the top 25% of white population, why not use median instead?
if only that data were readily available w/ like 4 clicks...
https://www.federalreserve.gov/econres/scf/dataviz/scf/chart/#series:Net_Worth;demographic:racecl4;population:1,2;units:median;range:1989,2019
This is very interesting, I'm very curious how this graph and Matt's graph can both be correct. I mean obviously they *can* be but it makes it seem like the precise distribution of black and white wealth deserve more scrutiny.
Or even ignore the top 25% of black people and then see what the numbers look like. I suspect the wealth gap would reappear 🤷🏿♂️
I think the proper way to illustrate this would be to take away the top 25% by wealth (regardless of race) and see what the racial disparity in wealth looks like in the bottom three quartiles. I assume it's much more pronounced than in the gap presented.
And yeah, focusing on the bottom three quartiles isn't indefensible, but, I mean, also someone at the 75th percentile of wealth is not particularly wealthy. If racial wealth disparities were entirely focused in the upper class, like if you got essentially no racial wealth disparity once you were looking at the bottom 95%, then sure, that makes the case that no amount of focus on racial wealth gap would meaningfully improve the lives of the overwhelming majority of black people. As-is, it does seem clear that a hypothetical program might reduce the wealth gap by specifically targeting black people in a way that affects the lives of tens of millions of everyday people.
None of that is to disagree with the fundamental point that just targeting programs that increased the wealth of the bottom 25%, or 50%, or 75%, would both a: reduce the racial wealth gap, b: be better politics, and c: help poor white people who do actually need help!
Apparently, the racial wealth gap is just as big percentage-wise at the median as at the mean:
https://www.federalreserve.gov/econres/notes/feds-notes/disparities-in-wealth-by-race-and-ethnicity-in-the-2019-survey-of-consumer-finances-20200928.htm
I think somewhere we got told "means are bad, medians are good," but medians can be misleading as well. Like if the deal is that 40% of white people have $0 wealth, and then it rapidly increases from there, and 50% of black people have $0 wealth, and then it rapidly increases from there, then median white wealth is plausibly $80k and median black wealth is $0 and it looks gigantic, but obscures that actually very similar percentages of both groups have $0 wealth. Not that this is necessarily actually what's happening! I don't know! But just contra "we can use medians for this."
Yes, exactly. This post is very similar to one by Matt Bruenig (https://www.peoplespolicyproject.org/2020/06/29/the-racial-wealth-gap-is-about-the-upper-classes/), and in both cases I disagreed with the analysis because they both assume that the mean wealth per group is the only way to measure racial wealth gaps.
Why is mean problematic in this use case? The size of the total pie matters here so even if outliers skew, that's a meaningful.
As the statistics joke goes, whenever Bill Gates walks into a bar, everyone there on average becomes a billionaire.
So ever since I read about this paper funded by Nick Hanauer I have just been gobsmacked by how much concentration of wealth at the top has been harmful to everyone else and caused most of the resentment and anxiety fueling a ton of our current social and political ills: https://www.rand.org/pubs/working_papers/WRA516-1.html If productivity growth had been shared more equitably from the mid 1970s to now all boats would be higher (they find median income would be doubled!) -- and hopefully we could focus more on eradicating racial and gender and other sources of inequality and discrimination. And the rich would still be plenty rich -- just not insanely so. I am super pro capitalism because markets do a good job of improving our material standard of living and pushing innovation, but we need to redistribute our GDP more equitably or we will never shake off all the conspiratorial crazy -- because there actually is a very small elite that controls an unhealthy amount of wealth.
"If productivity growth had been shared more equitably from the mid 1970s to now...And the rich would still be plenty rich -- just not insanely so."
So you didn't really mean "if productivity growth had been shared more equitably," you meant if more money had been taken from the most productive people and given to the less productive people.
Given that a big part of the spike in inequality arises from a declining income share for labor and an increasing share for capital, ridiculous to gloss reforms as "
"more money had taken from the most productive people and given to the less productive people".
The issue is that even as workers get more and more productive, the additional income from that output is going to owners (many of whom are passive, non-managerial owners).
“The issue is that even as workers get more and more productive, the additional income from that output is going to owners (many of whom are passive, non-managerial owners).”
If capital becomes more important why shouldn’t a higher share of returns go to investors? (And investing is never passive.)
I don't see how the concept of "capital becoming more important" makes any sense, analytically. All productive activity requires both capital and labor; they're both equally necessary and "important".
The relevant question is "how does the productive surplus get split between capital and labor", and that's seems like a question of economy-wide and society-wide structural factors, not some nebulous concept of "importance".
"All productive activity requires both capital and labor; they're both equally necessary and 'important."
Oh, you couldn't be more wrong. Imagine a line worker, in a factory, producing 20 $1 widgets an hour using a Widgetron-1000 that is owned by the firm that owns the factory. Further imagine that the worker's compensation (wages plus benefits) is $15 per hour. The firm's operating margin* is $5 an hour, so a 75/25 split labor to capital.
And then one day the firm buys a new-and-improved Widgetron-3000 that allows the same worker to produce 100 $1 widgets an hour.** The total output of the worker+machine (labor+capital) goes from $20 per hour to $100 per hour. Same input from the worker (1 hour) but five times the output for the worker+machine. This is clearly a situation where the labor and capital are not "equally necessary and 'important." Why, in your view, should we expect to see the same 75/25 split? (And before you answer, consider: If the returns to capital stayed the same why would the firm make the investment in the Widgetron-3000 in the first place?)
*This is obviously leaving out a lot of other production factors, from electricity to regulatory burden to maintenance to payroll taxes to opportunity costs. But, you know...widgets.
**Another oversimplification, of course, because the firm might instead decide to start selling the widgets for $0.50 each in order to increase market share.
Regardless of how productive the machine+worker combo is, the machine is worthless without the worker to operate it. The productivity of the two together defines the size of the surplus to be split between to two factors, but says nothing about what the division will be or should be.
The actual division of the surplus is a function of the leverage the two sides have vis each other, not marginal productivity--if, say, the factory unionizes shortly after the new machine is installed, the split that goes to the worker may well increase notwithstanding the vastly increased output facilitated by the machine; just as a prolonged weak labor market might drive down the worker's share even without any change in machines.
And although you're right that the firm wouldn't buy the new machine unless they thought they could extract enough of the surplus to cover the cost of the new equipment, that correct description of the firm's incentives doesn't describe how the surplus will or should be split once the new machine is in place.
"The actual division of the surplus is a function of the leverage the two sides have vis each other, not marginal productivity"
It is a function of that and a great many other factors. Indeed "leverage" is probably one of the lest important factors.
I take it, then, that your response to my question, "If capital becomes more important why shouldn’t a higher share of returns go to investors?" is, "Of course it should
What has the relative contribution of capital to output got to do with rent seeking?
B is not necessarily rent seeking.
What do you mean by "market structure"?
Well I mean that if the distribution of GDP across the income spectrum looked more like it did from 1945 to 1974 then we would have much higher wages for all but the top. I don't think that productive people were being shortchanged in that period of our history.
“I don't think that productive people were being shortchanged in that period of our history.”
Neither do I. Nor are they being shortchanged today. The reasons why “the distribution of GDP” (presumably you mean income) differs today are mostly pretty well understood: increasing returns to education, assortative mating and the resulting marriage patterns, etc.
"I don't use anything I learned in school at my job."
I highly recommend the book 'The Case Against Education' by economist Bryan Kaplan. Among other things he points out is that not only do most employees not use what they learned in school, but that most students don't even *retain* what they had learned. And he backs that up with data.
One of Kaplan's central points is that most valuable aspect of a college degree is not the education received but the credential. And that's pretty much what you seem to be saying about the "door in an elite world."
I don't know why you believe there wasn't money for elites 30 years ago.
"credentialism is a pretty obvious case of the market failing to accurately measure productivity"
It is quite the opposite of that. The credential is a pretty effective signal to a potential employer that a candidate will likely be productive.
"...both anecdotally and statistically, there's a lot more money sloshing around at the top"
Maybe we should eliminate the Clinton-era cap on the deductibility of executive compensation?
“It's tautological to say that: (1) productivity is defined by market outcomes, and (2) market outcomes are fair because they reflect productivity.”
Note that I did not say such a thing. Though it is true that market outcomes generally reflect productivity. (And, no, productivity is not defined by market outcomes. Productivity is defined as “the amount of output produced by a unit of input.” [Baumol/Blinder, 4th Edition])
“For example, non-competes are banned in California, which makes it easier for workers to switch jobs and tends to make market outcomes more equitable.”
How do you know? One certainly shouldn’t draw that conclusion based on California’s above average Gini coefficient: https://en.wikipedia.org/wiki/List_of_U.S._states_by_Gini_coefficient
"That only makes sense if you're measuring productivity based exclusively on market outcomes"
In a market economy how else would one make that measurement?
"...your argument goes: (1) productivity is reasonably reliably measured by market outcomes"
I cannot think of a better one. Can you?
"your argument goes:...(2) productivity is a fair way to distribute wealth, therefore (3) market outcomes are fair."
I did not say that nor do I think that.
"But we've seen decades of rising income inequality"
Perhaps. So what?
"you said that reducing income inequality means taking from 'the most productive people'"
You imagined that.
(2) is also pretty contestable! Given what we know about diminishing marginal utility of income, there's a strong case that we'd be better off doing more redistribution than we currently do.
Many (including me) agree that most people aren't rich. But what you don't address is that the history of universalistic programs is rivened by white racism. The history of social security is well known, But even the most recent attempt to benefit everyone--the Affordable Care Act--was opposed by white people everywhere, especially at first. It required a supermajority in the Senate and a Black president to be enacted--and even so Tea Partyists used its passage to elect an ACA opponent in MA. In short, what do you propose to do about white people who oppose universal programs mainly because they think universal programs benefit non-white people? Asking for me.
"In short, what do you propose to do about white people who oppose universal programs mainly because they think universal programs benefit non-white people?"
They are exceedingly few in number, so I suggest ignoring them.
If they are few, they are exceedingly well organized and/or determined. They have managed to stymie universalist programs since at least Truman.
Where is the evidence that they have done so?
As long as some percentage of opponents can be turned into supporters by focusing on class, we outvote the rest and wait for their (likely less racist) children and grandchildren to start voting.
As someone mentioned in yesterday's discussion, election victories and the policies they produce are built on small margins.
So I’m going to preface this with the acknowledgment that in the 90s Clintoncare similarly failed to achieve support for a universal program despite not being pushed by a white administration (granted with Clintoncare you had to dual problem of Hilary and sexism as well as the very very racialized crime reform driven by the very real crime problems of the late 80s/early 90s).
(And further caveat: I wasn’t really alive for all this, so my knowledge is all filtered through reading about people reflecting on this more than 20 years later)
So all that being said, I’m actually very curious about the counter factual scenario of a boring milquetoast centrist white democrat proposing Obamacare in the similar landscape of post 08 crash and huge democratic majorities? What happens then?
Could the Republican leadership (even if they were still determined to oppose) have held their caucus?
Maybe, but I doubt it because the base motivated some of the solidarity.
Would that crazy base that said they hated socialism even if it was pretty obvious they were motivated by racial grievance have come out? (For proof: see how they loved Sarah Palin’s true American schtick in fall 08 and the seeds of birtherism already coming out that fall)
Of course, say I am right, what to do now? Because this posits a very dark picture of a significant portion of the American electorate which doesn’t portend great things for us as a multi-ethnic democracy. How do we solve the problem of fear of the other, which is almost exclusively coded as black or brown? And Idk. It’s a pickle.
"So I’m going to preface this with the acknowledgment that in the 90s Clintoncare similarly failed to achieve support for a universal program despite not being pushed by a white administration (granted with Clintoncare you had to dual problem of Hilary and sexism as well as the very very racialized crime reform driven by the very real crime problems of the late 80s/early 90s)."
Just as opposition to Obamacare wasn't about race, opposition to the Clintons' attempt at health care "reform" wasn't about sexism. Rather, opposition was primarily about resistance to handing over so much power to a massive new bureaucracy and requiring employers to provide health insurance for all employees. (Let me know if you are unclear about why both of those were - and are - terrible ideas and I'll fill you in.)
Also: Look up "Harry and Louise" on YouTube.
Of course it was about race, at least in part lol. Obama being depicted as a witch doctor wasn't some high minded statement about government bureaucracy. It's a racial anxiety trigger.
I don't have strong views on Clintoncare or whatever, I prefaced with it only to head off it the argument that Republican opposition would have been as united in 2009 as the 90s even if Obama hadn't been there to stoke the racial anxieties of white americans. My argument is there is a good chance that it would have been different. Because 2009 was not the early 90s. The political realities were different and without the racialized opposition coming from the base, the republican caucus might not have held
"Of course it was about race, at least in part"
An insignificant part.
This is being done by 40% (or less) of the country that Democrats will never get, no matter what they do. We have a bad habit of focusing on them, as if they represent everyone who voted Republican. The 8% who got Trump.from 40-48%? These are the voters who matter. And they're not birthers.
"The history of social security is well known, But even the most recent attempt to benefit everyone--the Affordable Care Act--was opposed by white people everywhere, especially at first."
Not everywhere. The Affordable Care Act was essentially a nationwide version of the Massachusetts program Mitt Romney signed into law in 2006, which seems to have been well-received by white people at the time: https://www.theatlantic.com/politics/archive/2011/12/newt-gingrich-celebrated-romneycare-in-2006/250531/
True enough. So then what do you think accounts for the election victory of Scott Brown?
I can only guess at this point, but it looks to me like there were a lot of issues in play during that election, including some missteps by Coakley. As Barack Obama said in an interview with New York Magazine:
"Well, the first thing that’s happening is I’m talking to Rahm [Emanuel, then chief of staff], and Jim Messina [then deputy chief of staff] and saying, “Okay, explain to me how this happened.” It was at that point that I learned that our candidate, Martha Coakley, had asked, rhetorically, “What should I do, stand in front of Fenway and shake hands with voters?,” and we figured that wasn’t a good bellwether of how things might go."
This Boston Review analysis points to both lowered Democratic turnout and an erosion of support among Dem-leaning minority voters: http://bostonreview.net/charles-stewart-stephen-ansolabehere-what-happened-in-massachusetts-scott-brown-special-election
It seems to me that programs like public education, Social Security, Medicare and Medicaid have been extremely beneficial for black and white people. I'm not sure if I'm going to do anything to people that disagree with me regarding programs, however, but continue to advocate for ideas that make a ton of sense.
Unless I've totally misread the take here, what this (very good and informative, by the way) post represents is the intersections of Matt Yglesias and Bernie Sanders.
...and maybe also John McWhorter
https://www.theatlantic.com/author/john-mcwhorter/
I think taking off the top 25% is statistically silly. If a baseball batter is one for four on average, he's mediocre. Make him one for three, and he's one of the best hitters in the game.
If you don't like that part, then just look at the part about how the top 10% is whiter than the general population, the top 1% is whiter than that, and the Forbes 400 list is even whiter.
The more elite the economic group you look at, the whiter it is. That's a true fact and it impacts the statistical aggregates. But diversifying the billionaire class would do nothing to help the typical Black family.
I agree that somehow "manufacturing" more black elites is performative, not substantive. To me, mobility is the key..
Raj Chetty has shown the racial disparity in mobility across generations. To your point about land use, Chetty has also shown a very positive correlation between mobility and time spent growing up in a more socio-economic diverse zip code. That's more important than almost any other factor. So affordable housing in your MA town example would be exactly on point
TL; DR Please use histograms when discussing distributional issues.
David Roberts has a point that lopping off the top 25 percent of the distribution is silly. That said, Kevin Carney (and by extension you) have good reason for wanting to do it; to only show the means (medians) does not tell you a whole lot about how two distributions compare because you are only reporting one moment (quantile) of each distribution. However, the better solution would be to show ALL of BOTH distributions (i.e., by using a histogram or nonparametric estimate of the density), like Pew Social Trends does with income here: https://www.pewsocialtrends.org/2018/07/12/appendix-a-income-distributions-of-whites-blacks-hispanics-and-asians-in-the-u-s-1970-and-2016/.
The only downside is that you need more figures (and therefore more space) to show how the comparison changes over time. In this case, however, it seems like the emphasis of the story is the comparison of the two distributions, not the time trend of that comparison, so this shouldn't be a big deal.
Yes - a fuller comparison would be to compare the distributions of each curve.
It also wouldn't help the typical white family as well. This is one of the things that conservatives do so well when they show their black politicians or supporters and make white voters feel good without actually doing anything for them. It creates this veneer of diversity while at the same time letting the politicians play on racial resentment to win political races. If poor white voters were instead united with poor black voters in fighting for things that actually improved their lives you could have a very powerful political party. Poor white people feel that they are connected to rich white people in some weird convoluted way even though they are constantly getting taken advantage of (i.e. Trump's election defense fund that was just used to pay off campaign debts). Doing the same thing for black people by making them somehow connected with black elites plays on that same racial cord that Leftists criticize conservatives using and does nothing to actually solve underlying problems (it actually makes them worse i.e. Obama winning the White House and the backlash that ensued)
Sure but poor white people are still very closely connected with race. It's not like they are uniting with poor black people to form a big coalition against white, wealthy people. They are still voting for conservative policies that keep white, wealthy people in power (i.e. Donald Trump) even though it goes against what would actually help them (such as Medicare and welfare). So I see your point that they may not have a connection with the new wealthy white people but they still believe they are better than poor black people.
This is the C Vann Woodward thesis and the problem is that he wrote that in 1955. I question whether there is not something else going on in the more than half a century since then. What I see is an egalitarian culture gone rancid. Trump people hate liberals, not blacks.
This may all be true, but racial wealth gaps as measured by median wealth are persistent across income level, education level, age, etc.
Even the pew research that you cited demonstrates these persistent gaps in median wealth.
I tend to agree cash payments are unworkable. But I'm sympathetic to more targeted measures that address more narrow systemic examples. So here in Chicago I would support a program that increased or even re-distributed school funding to previously redlined neighborhoods in proportional amount to cumulative loss in tax revenue from limiting the capital appreciation of the housing assets. But maybe Chicago was so egregious here in their application of redlining that I'm biased vs. the national view.
I think Matt would argue that a simple formula change to allow for more funding in poorer districts (which is in many cases what we have actually done) would be easier to justify to voters and accomplish the same thing.
Hopefully Matt weighs in. I'd be interested in his take. Chicago in so many respects is simply inept.
That sounds like an incomplete analysis. Giving poor people money lets them spend it, so it would improve the life of anyone who can sell things to them.
I didn't interpret it as anything statistically rigorous, I think it just illustrated that the economic experience of the vast majority of white Americans is very similar to the economic experience of Black Americans. I find it a useful framing to keep in mind.
I think that it’s worse than that. There is an abundance of empirical data that demonstrates persistent racial wealth gaps measures by median wealth. The statistical trick employed uses mean wealth, so the ultra wealthy black population’s wealth is included while that of the white population is not. It doesn’t compare apples to apples.
Comparing median wealth would be more representative of “typical experience”, which they did not do.
The pew research that Matt linked shows persistent racial wealth gaps across income and education levels!
The homeownership rates are persistently higher for white households, which supports the story of a broad based wealth gap and agrees with the documented history of government policies that made it easy for white families to get a federally subsidized mortgage while preventing black families from doing the same for decades.
It's still relevant though because at the end of the day we are still talking about individuals. Your numbers are important, too, but the fact that you can find huge numbers of white people in the same situation as analogous black people does imply that race may not be the best thing to focus on in order to enact meaningful change.
I'm confused about the mean/median difference. Given the right-skewed distribution of wealth in the US, I would expect the gap in Black/white median family wealth to be smaller than the gap between their wealth at the mean. But according to a Fed publication, that's not true: Black family wealth is 15% of white family wealth at both the mean and the median.
Source: https://www.federalreserve.gov/econres/notes/feds-notes/disparities-in-wealth-by-race-and-ethnicity-in-the-2019-survey-of-consumer-finances-20200928.htm
My point about focusing on medians is that this is a simple way to negate the skewing effect of outliers, which is presumably what the impetus behind lopping off the top 25% of white wealth was. It's a weird thing to do.
The point of looking at medians is to compare the 50th percentile of white families to the 50th percentile of Black families. The fact the the gap persists is precisely the point that I was making - the racial wealth gap is not just the result of there being more super-rich white billionaires.
To your specific point: the link you provided shows a slightly larger gap for when using medians - Black median wealth is 12.8% of white median wealth, while Black mean wealth is 14.5% of white mean wealth. What I take from this is that wealth is right-skewed both among Black families and white families.
I agree with you! My confusion was that I expected there to be a much bigger bias when using means instead of medians. And that turns out not to be the case.
I also found it to be an interesting result! I hadn't seen this before, thanks for sharing!
Just want to echo that you're right, RS. Use of median in skewed distributions is 101 stuff. Love Matt, find the political tactics here reasonable to consider, but this post's premise is innumerate at the core
https://www.brookings.edu/blog/up-front/2020/02/27/examining-the-black-white-wealth-gap/
That's not analogous, though. A baseball batter is a single person, and the average is a single number intended to summarize his ability. Looking at individual hits by themself would actually be deceptive, since you want to know their performance long term.
The case of a group's wealth is more complex than that. "White people" are a group of individuals, and what we really care about are the individuals, not the aggregate. For a white person in the bottom 25%, the wealth of the top 25% does them no good, and a statistic like the one Matt presented helps illustrate that point.
Another way to think about it is this: dividing by race is inherently arbitrary, you could just as easily divide by hair color, or favorite pizza toppings, and the distributions would still be valid; "the bottom 75% percent of white people" is potentially just as valid a group as any other, as long as you contextualize why you're using that metric. On the other hand, there's no alternative way to divide a baseball player's hitting average.
And this is why you Matt had to leave Vox
CSB: I used to be a union organizer. One of the (many) things that we had to fight was "bosses"/conservatives driving a racial wedge between people who were fighting the same fight. And now " liberals" do it; if you even attempt to explain why this is a bad idea, they will call you racist.
Like many things that can seem incidental or coincidental, this isn't. The kind of people who fund the CAP (for instance) would much rather hire three black women and have everyone tell them how awesome they are, than give 300 people a raise (including *30* and genuinely make a
(Including *30* black women) and genuinely make a real dent in the problem. (Whoops, I hit post too soon, sorry)
Definitely an interesting read.
All of the "just lop off the richest quarter of white people" analysis seems very suspect to me. Excluding rich whites and not excluding rich Blacks would be expected to make a wealth gap disappear, right?
What happens if you re-run all this analysis after also lopping off the richest quarter of Blacks. I suspect that you'll find a racial wealth gap again.
I'm not going to look this next bit up to verify it (poor form on my part, but it's very early in the day), but I remember reading once that the median net worth of Black families in Boston was effectively zero. The median wealth of white Boston families is presumably not zero.
Idk man. I think sometimes people try and reduce everything to class because "bring in the tide to lift all boats" is an easier sell than "bring in the tide but also plug the holes that people sawed into the hulls of 12% of all boats".
I'm always glad to read what you've written though, especially when we disagree.
Page 21 of the paper Matt linked has a graph of the wealth gap between the median white household and the median black household. It shows median white household wealth bouncing between $60k and $100k 1989-2013, with an upward trajectory for the 90s and 00s offset by a big fall with the Great Recession, ending up at $80k. Meanwhile, median black household wealth started the period at $10k, rose to nearly $20k in 1998, but declined every year after to almost $0 in 2013.
I think you're on to something but the solution would not quite be the one you suggest.
You could look at the wealth level of the 25% percentile (for example) of whites, and then "lop off" anyone (regardless of race) above that level from the analysis. This would be the proper way to investigate this question with the, um, "lop off" mechanic.
If Matt is correct, this would almost entirely remove rich white people from the data and his point would hold. In theory though, compared to the analysis he describes, it could instead remove enough rich black people from the data that the racial wealth gap would be evident.
(I believe the other evidence for his point is strong, to be clear, but I do agree that something about this particular point of analysis seemed a little off.)
I don't disagree with what you've said, but at the end of the day if all you do is focus on trying to lift all boats then (pending someone doing the analysis) you aren't going to get rid of (all) the wealth gap.
If that isn't your highest priority then no problem (it isn't mine either!), but that's different than chalking up the gap to just being a class gap in disguise.
I think some would argue that you are sort of side stepping the institutional racism argument. That argument would suggest there is an extra barrier on top of class that will still make it difficult for some people to rise economically even if you try to adjust some of the prosperity metrics through technocratic innovation. Do you have an opinion on how ingrained racism could be overcome directly rather than circumvented?
I mean as I said at the end, I think you need to tackle racial discrimination as an autonomous sphere of activity. If academics can do blind resume tests to detect labor market discrimination, then the Civil Rights Division of the Justice Department can use the same tactics to do stings.
We have some workable models of police reform. We can continue the process of promoting more diverse representation in political office.
In other words, I think most more or less "normal" ideas for fighting discrimination and racism make sense. But attempting to directly target the racial economic distribution is a confusing idea, given the underlying circumstances of severe inequality.
I see where you're coming from, but this sounds a bit too authoritarian for my taste.
Also, depending on organizational goals and culture, being able to identify gender and racial cues on a resume may actually increase hiring of minorities.
This is anecdotal, but I ran an engineering organization and we were always behind our diversity goals. I'm a bit embarrassed to admit it, but if I'm being honest, whenever I saw a resume with a women's name on it, my filter changed from "is this person an outstanding top of their class engineer" to "is there a way we could possibly make this person work in our organization."
A French government study in 2009-10 found the same thing, which led to the government abandoning that policy in 2015: https://www.povertyactionlab.org/case-study/unintended-effects-anonymous-resumes
"When resumes were made anonymous, participating firms became less likely to interview and hire minority candidates. The gap in interview rates between non-minority and minority candidates widened by 10.7 percentage points, from 2.4 percentage points in the standard procedure to 13 percentage points in the anonymized procedure. At the hiring stage, the gap widened by 3.7 percentage points."
Interestingly, because the study was opt-in on the part of employers, the researchers' conclusion was that the companies eager to adopt a new diversity program tended to be the ones whose biases ran in the opposite direction to begin with:
"Why did anonymized resumes lead to fewer interview offers to minority job candidates? The answer may lie in the difference between firms that selected to participate in the program and those that did not. Participating firms that were randomly assigned to the comparison group (and thus continued to receive resumes with all identifying information) were twice as likely to offer interviews to minority candidates. The average firm volunteering to participate in the study thus appeared to positively value minority candidates’ resumes in general, and anonymization changed the decision-making process for these firms."
My take is that imposing a legal bureaucracy in charge of standardizing resume inputs is not going to get us what we want. But we can and should definitely advocate that companies or hiring managers do that.
To offer a personal story, I recently hired from a pool of about 50 applicants and had an excel file to pick who I wanted to interview from. I hid the applicant names and university names and only evaluated each candidate based on their credentials like degrees, certificates, and listed technical skills or knowledge.
From this "blind" process, I picked 7 people to interview: 1 black female, 1 Asian female, and 5 white men. I ended up hiring of the white men who was clearly the best candidate, but I was pleased that my little process seemed to have generated an equal-opportunity process.
And they've been doing it for a while in orchestras and now are backtracking because it didn't do what they hoped: https://www.nytimes.com/2020/07/16/arts/music/blind-auditions-orchestras-race.html
Right. Because who cares about those female players who were helped by the policy.
My understanding of the classic blind-auditions study is that actually the original data analysis was bad and it was a null result, and no one has ever produced a successful replica study.
Female orchestra players increased significantly after the policy was implemented. Occam's Razor suggests that the policy was the reason for the increase. I think the argument that audition panels suddenly had a change of heart is quite a bit less plausible.
One way to test this presumption would be to conduct a statistical analysis of the effects of the policy change on a per-audition basis, adjusting for changes in the relevant populations. Turns out that, when you do this, you don't find any substantial evidence that the policy improved the outcomes of female musicians, and some evidence that it may have hurt their chances.
The advantage of doing this, rather than reaching for Occam's Razor, is that it helps to disentangle the effects of a single policy from the more general societal changes which have boosted female employment outcomes in most industries over the last fifty years without the adoption of blind auditions.
"general societal changes": As I stated earlier, I am skeptical that this was the cause of the change in female representation.
I find it interesting that a number of people here are pretty invested in the idea that blind auditions don't help women.
I think instead we just need to make those traits not as valuable. We need to change the biases people hold rather than forcing bias to not be in the process. How would you prevent gender from being an issue in an interview? Would you just black out the person's face and distort their voice? We need to change the stereotypes we hold to be more productive in society rather than trying to get rid of stereotypes altogether. I'm not sure exactly what types of stereotypes would be most beneficial but I do know that trying to remove a crucial part of the human brain is a very difficult and arguably impossible road to go down.
I pitched a seed round start up for this. The idea was a CarFax like solution for HR departments. It would autonomize resumes prior to submittal. I still think it's a good idea.
They tried this in some tech companies but it didn’t benefit women. https://blog.interviewing.io/we-built-voice-modulation-to-mask-gender-in-technical-interviews-heres-what-happened/
White guilt isn't completely worthless but you are right that it isn't the best way to get things done. I think this is one of those "two things can be true" issues where we can probably find sensible ways to address both issues but it would take the kind of consistency of message and coordination that the current Democratic Party lacks. Maybe Matt is purposefully circumventing that part of the argument because he realizes it is divisive and he is trying to cut through to a solution that could help in a very utilitarian way.
Bo can you share the value of "white guilt?" I've seen a lot of evidence and discussion about how attempting to shame is counterproductive. Would be curious how you see this being an exception? Especially if it inclines poor whites to object to economic redistribution?
In the "value" order I certainly don't put it above lots of other useful motivations. I do think a small amount of guilt or shame for things in our past can be useful as long as it is properly contextualized and the resulting emotion is put toward accomplishing good in the world. I do not think it should hoisted on a brand banner for the whites on the left to rally around, I think it has a small personal motivating use for some people. That's really the extent of it. So not useless but also not something that is useful to champion.
But the "makes them feel good" is the inherent moral imperative that comes along with a lot of stuff we do in our communities. We like to do stuff that we feel has real social benefit, that makes us feel good. I agree that the hazard is when this becomes a form of selfishness. Some people may be frustrated because they feel like lots of things have been tried already and they are not seeing results. We just need to be mindful of that headspace and respectful of the difference in peoples intentions.
I am 100% for less sanctimoniousness and more empiricism : )
It may be a commonplace now, but I’ll say that I live in a mostly white, professional neighborhood where you can see lots of BLM signs - and also lots of people who are ferocious opponents of more-dense housing or the housing voucher policies that would open the neighborhood to less affluent residents.
Good piece. I think you can make the same point about university admissions. I don't get why there are so many battles over racial diversity in college admissions. To me it makes more sense for colleges to simply allot more spots to lower income families and that should increase diversity along with it. Unless they don't really care about assisting lower income kids and would be happier taking in upper class minorities.
I think this ties into another, related question, one which maybe Matt would like to ponder in another post: why has the US -- pretty much alone among Western nations -- never had a strong, let alone successful, socialist movement?
Many have attributed it to an ingrained suspicion on strong governments. Probably true, but also true is that most Western nations have historically been fairly uniform ethnically, racially and mostly religiously; their major dividing lines have been on class lines. In the US, class never has been the major activating feature of group conflict, whereas racial and ethnic divisions have been. Thus the default tendency in the US -- both on the left and the right -- is to see things divided on racial and not class lines.
If that's deeply embedded in our culture, I suspect class-based policy initiatives may have trouble succeeding.
I think people are massively overstating the importance of naming conventions. The British Labour Party is technically a socialist party but Tony Blair and Gordon Brown saw Bill Clinton as a kindred spirit. Indeed, long before him, a Labour Leader said that "English socialism owed more to Methodism than to Marxism". All Europe's key socialist parties ultimately renounced Marxism, most decades before the end of the Cold War. America has a history of powerful trade unions, it has a history of public agitation for key utilities to provide through government funding, it has a left wing party that is closely aligned with trade unions, an alliance that has endured as the centre of gravity amongst those unions moved from heavy industry to the public sector.
Ultimately these arguments come down to "why did the American left not adopt the naming conventions inspired by arguments of a German economist who lived in London", I dunno, maybe because America is half the world away from either country. There's no American Tory Party either, but no one goes round saying that America is unusually left-wing.
Interesting but not really related to what I was saying. And that was that the primary cleavages in American politics have been on racial and ethnic lines and not class lines. I don't think you can say that about the Labour and Tory parties in England historically.
Actually there's a lot of demographic divides between the Tories and Labour/Liberals (basically English vs Celts, Anglicans vs Other Protestants, South vs North, Rural then Rural/Suburban vs Urban), and even between Labour/Liberals (Catholics vs Non-Anglican Protestants). Likewise Britain's BAME community has been a solid Labour voting block, not as much as African-Americans, but comparable to (say) Hispanic Americans. The actual difference is that the brief periods progressives get control in America, they can't push through as much radical change as British ones can, because our system has virtually no veto points. But that's also why our Thatcherites were able to privatise the buses, trains, air traffice control, post office, etc whereas Reganites couldn't.
I think Bernie was the most successful socialist movement, but I think it is the type of socialism that became popular in Western and (especially) Northern Europe: social democracy. I think its movement is stuck at the moment due to running poor campaigns, but I would be happy if a social democracy movement were to start to gain steam
This is an interesting theory.
I wonder if our increased income inequality (concentration of wealth with fewer people) will increase class dividing lines.
I'm pretty conservative, but I dislike wealthy people. If was king, I would tax Gates, Buffet, Zuckerberg, etc... so much they would be driving uber to pay bills.
Or at the very least, a 90% death tax on all estates values above 50 million.
I'm pretty pro founder billionaires - or certainly more so than your typical progressive. I think A LOT of the founder billionaires are simply the extreme end of the distribution. It's a biased sample. You're not measuring the millions of failed founders and I think there's a reason Silicon Valley exists here and not elsewhere and favorable tax structure plays some part.
WHAT I CAN'T AT ALL FATHOM is how much offshore tax dodging we allow and why we can't attack inter-generation wealth transfer. To me it's all criminal. It's hurts to see all the Walton heirs atop the Forbes list. I think Sam built an amazing business. I don't think his kids did anything other than win a lottery.
Agree. There's no greater equal opportunity cause than dramatically hiking inheritance taxes. Let founders who genuinely created something enjoy their wealth while they're alive, but when they die distribute it equally to everyone, not just a few undeserving people who won a parental lottery, and then think got it because they deserve it.
I am pro founder billionaires as well... ok... Im technically pro take a lot of their money while they are alive via wealth tax and take the rest when they die via death tax. And Im not even liberal. I just dislike rich people.
The US has had some very strong *socialists* (maybe not TR!) but not a strong socialist *movement.* And, yes, I define that as winning elections, implementing socialist ideology or maybe leading popular movements to take over government through other means.
Indeed, FDR incorporated thinking of some socialists. So did Bismarck under Staatssozialismus. Neither, obviously, was a socialist.