Ban The Box & Racial Discrimination

Employer inquiries into the criminal background of job applicants could actually lessen racial discrimination against minority applicants. That’s the conclusion of a much reported on field study by a pair Michigan economists. Using the adoption of Ban The Box legislation in New York and New Jersey, the authors reported a six-fold increase in the disparity of callbacks received by white and similar minority prospective employees after the legislation took hold. Without recourse to an applicant’s criminal history, employers seem to be assuming that minority applicants have criminal backgrounds.

I find it unsurprising that establishment media outlets have hailed these results as “fascinating” and “unsettling“. Its an alleged poster child for the failure to consider unintended consequences, and it fuels the narrative that societal problems require intricate, non-intuitive solutions. Sendhil Mullainathan summarizes this ethos well: “When we try to end discrimination without addressing the underlying causes of discriminatory behavior, our efforts may accomplish little — and may even backfire.” Some commentators appear to suggest scrapping Ban The Box legislation altogether.

Missing from the liberal response to these findings is Ban The Box’s impact on Title VII prohibitions against racial discrimination and the use of criminal background checks to keep the arrested and convicted unemployed. Proving racial discrimination in the employment context is a tall order, with those alleging unfair treatment likely relying on indirect circumstantial evidence. The Supreme Court established the framework for proving indirect charges of discrimination four decades ago. Once the plaintiff shows that they are a racial minority and were not hired despite adequate qualifications, the burden shifts to the employer to offer a nondiscriminatory reason for their decision to not hire. Employers can and do satisfy that burden with HR policies forbidding the hiring of applicants with criminal records. Criminal records and race covary, leaving minority job-seeekers unduly exposed to the employer’s right to discriminate against criminals. This reality persists despite the EEOC’s increased vigilance towards blanket policies against applicants with records.

Ban The Box removes a powerful tool from employer’s arsenal. When a minority applicant brings a lawsuit alleging racial discrimination, employers will have to concoct a story that doesn’t depend on a racially-biased criminal justice system. Those touting the policy’s unintended consequences seem to discount the significance of employer’s apparently unlawful response. There are few clearer examples of an employer’s discriminatory decision-making than the suspicion that all minority applicants have criminal backgrounds. Ban The Box exposes employers to liability for racial discrimination while protecting the victims of an expansive criminal justice system.

It is mostly true that the law cannot change people’s attitudes and prejudices. Acknowledging this, our politics should be steeped in a commitment to minimize the power of people to act on those prejudices. The coverage of recent studies reporting on the unintended consequences of Ban The Box initiatives, I think, stems from that understanding. It’s been my attempt here to show why the trade-off between discriminating against minorities and against people with criminal records is a choice of under-enforcement of the law and not necessity; why Ban The Box could kill the two proverbially birds with one stone. Before letting you go, though, I can’t help but mention a certain prejudice present in reporting on this topic – people seem a bit too willing to accept and justify the exclusion of people with criminal histories from the workplace. That’s something to think about when we evaluate a policy that would categorically prevent it.

Property Rights & Production

Reading Demsetz’s “Toward a Theory of Property Rights” for class. Demsetz’s thesis is uncontroversial and, I find, a helpful analytical principle for understanding long-run changes in property rights. Property rights, he writes, develop to internalize externalities when the gains of internalization become larger than the cost of internalization. Or, in other words, private actors end up shouldering more of the costs of an activity when the costs to society of the private actor’s activity outweigh its benefits.

His thesis gains evidentiary support from the correlation between property rights in land and the commercial fur trade among American Indians. Absent property rights in land, hunters have no reason to moderate their hunting to a level that sustains animal populations over time. This is particularly so when the price outsiders are willing to exchange for furs is rising. To stave off the threat of resource depletion, hunters begin marking off land as “theirs”, and property rights in land are born*.

This story is one of the most popular defenses of property rights in land. In contemporary microeconomics, it gained credence under the name “Tragedy of the Commons”. The story Demsetz tells is difficult to refute on its assumptions. But, like much else in Econ 101, its assumptions are tenuous. Most glaring is its starting point: a society where laborers have a property right in the fruit of their labor. This personal property right is what incentives additional hunting activity in an era of rising prices. Where hunters have a property right in the animals they kill, property rights in land internalize the external cost of over-production.

Where there is no threat of over-production – for instance, where hunters do not have a property right in the animals they kill – property rights in land serve only to restrict the freedoms of non-landowners. Per Demsetz then, the property right would not develop. The decision to hunt falls upon some other entity that requires hunting as a productive activity rather than the private actor that views hunting as a means of personal enrichment. Perhaps a group of people dependent on hunting gather to make decisions about appropriate hunting levels and distributions. Perhaps they delegate that decision to a third-party in return for a legal claim on a portion of the animals hunted. The ways this could go down are many. I generally dislike meetings, so I would probably vote to delegate the responsibility.

It could be countered that hunting would grind to a halt where hunters do not have a property right in the animals they kill. To this, the experience of laborers in the United States is instructive. Despite having no enforceable right to the goods they produce, American workers regularly exchange the value of their labor for a 30-40% markdown that they can then exchange for means of survival in consumption markets. American capitalists then, having both a legal right to property and legal right to the fruits of labor expended upon that property, are tasked with internalizing the threat of over-production. Their record on this is decidedly negative.

There’s no grand conclusion here, not yet at least. Just a few observations.

*It’s unclear to me why the community that cannot police its members to not over-hunt can police its members to respect others property rights in land.

I am 76% sure that the August jobs numbers will be revised up

At 8:30AM on the first Friday of every month the Bureau of Labor Statistics (BLS) will release their initial estimate of employment growth in the preceding month. By 9:00AM on the first Friday of every month a handful of well-established economics’ reporters will move one step closer towards their weekly quota with a two-part piece. They’ll start with a comparison of the BLS’s initial estimate to the projections. Then, as a bit of insider knowledge, they’ll remind us not to read too closely into those numbers; in three months, after all, the BLS will release more accurate, revised figures.

It used to be that Dean Baker was the only person making reference to the revisions. Now, you can’t write anything without granting them at least a paragraph. That’s definitely a good thing. Thank you Mr. Baker.

In the general interest of having something new to read every so often, I put together the visual below. It is a graphic representation of the last five years of monthly BLS reports. When the BLS releases a initial estimate, they include a 90% confidence interval. Usually it is +/- 100,000 jobs, though it varies annually. So, last Friday we learned that the economy added 173,000 jobs in August. Using the confidence interval (105,000 for 2015), the BLS is then 90% confident that the actual number of new jobs is between 68,000 and 278,000. The black bars attached to each initial estimate (red square) represent the breadth of the BLS’s confidence interval. The blue diamonds show the revised (and more accurate) figure. Because we are not yet three months past July or August, the graph does not include those revisions.

Two things that I immediately noticed. First, on eight occasions (out of 66) the revised estimate came in above the figure that the BLS had 95% confidence the actual job growth would come in below. The odds of this happening: 1.145 out of 100. Unlikely, but possible. Second, in 50 of the 66 months, the BLS revisions came in above the initial estimate. Revisions upward should happen as often as those downward. Instead, over the last five years initial estimates have risen upon revision 76% of the time. The odds of this happening: 1 in 86,258.

It looks like there’s a flaw in how the BLS is surveying and interpreting data. Economic commentariat, what is it?

Social Security’s ‘Solvency’ Is Irrelevant

This year marks the 80th anniversary of the U.S.’s most successful anti-poverty program. In a world where freakishly intelligent people devote much of their existence to building innovative social programs – think charter schools, opportunity zones, inclusive zoning – it’s really quite embarrassing how simple the standard-bearer for poverty alleviation is. By simply mailing a monthly check for ~$1180 to residents 65 years and up (with some exceptions), the Social Security Administration is able to lift 25 million people above the poverty income threshold every year.

The rhetoric around Social Security has ignored it’s success as an anti-poverty program for some time. Instead, politicians of all stripes worry publicly about the program’s future solvency. Their hysterics, perhaps unsurprisingly, mirror the concerns of the affluent, and they seem to have conjured great concern among younger people. Despite Social Security being wildly popular, 51% of respondents to a recent Gallup poll doubt that they will get benefits upon retirement. That’s a shockingly high number for a pay-as-you-go program, and it makes you doubt whether people really understand how Social Security works. As I’ll explain below, the only way you won’t get Social Security benefits is if Social Security is scrapped entirely or is fundamentally altered as a universal program. Short of that, you are guaranteed benefits.

Social Security, unlike most other social programs, is entirely self-funded. For every dollar your employer pays you, s/he sends $0.062* to the Social Security Administration (SSA). In addition, for every dollar you receive from your employer (or at least the first ~$118,000), you send another $0.062 to the SSA. So, in good times and bad, the SSA collects 12.4%** of the nation’s payroll to operationalize the nation’s largest poverty program. They receive no funding outside of the payroll tax***. When Congress completes their annual ritual of drafting a budget to be sidelined in committee, exactly zero of those budget dollars are earmarked for Social Security.

The SSA collects their funding from the 12.4% payroll tax and uses it to finance retiree benefits in accordance with a formula Congress legislated. This formula yields an average benefit of $1180 per month per beneficiary. Some years, they collect more in taxes than they distribute. Other years, they spend more than they collect. When SSA collects more, they run a surplus and the surplus is placed in the Social Security Trust Fund and invested in non-marketable securities. Conversely, when they spend more, they sell off assets in the Trust fund to finance the deficit.

For a long time, the SSA ran surpluses and built up a sizable Trust Fund. Assuming a constant return on investment, there are two ways that this can happen. One, the workers:retirees ratio grows. More workers relative to non-workers means more people to tax relative to people to spend on. Two, economic growth. Holding the worker:retiree ratio constant, a higher growth rate means a larger payroll for the 12.4% tax to be levied on – it also questions our prior assumption of constant investment returns. The postwar economic boom and the entrance of the Baby Boomers into the workforce delivered on both counts, and the Trust Fund grew.

Slowing economic growth and the retirement of the Baby Boomers has combined to weaken the outlook for Social Security’s Trust Fund. Recent estimates suggest that the Trust Fund’s asset/cost ratio will approach zero in twenty years. That means that if in twenty years the SSA collects less in taxes than they pay out to beneficiaries, there will be no assets to sell to cover the difference. What happens then?

Well, the SSA estimates a 21% shortfall, meaning they’ll be able to pay to beneficiaries $0.79 of every dollar their formula says is owed. This would be a disaster to Social Security’s standing as the nation’s most successful anti-poverty program. However, it wouldn’t spell a disaster for Social Security. So long as 12.4% of payroll is collected, at least 12.4% of payroll can be spent. Cutting Social Security benefits now to save it later misses this crucial point. You can’t get blood out of a stone; 12.4% is 12.4%. The program’s ‘solvency’ is irrelevant, a silly lens to analyze a program through that is forbidden from borrowing. Put another way, what good does it do to cut benefits today because you are worried about funding being short tomorrow? Either way, you are collecting and spending 12.4% of all annually created value.

It’s quite possible, and in my opinion very likely, that 12.4% of all value created will be insufficient to provide future retirees with an existence that isn’t also impoverished, but that’s a wholly different discussion. It demands attention to income inequality and requires the use of a theory of just distribution to determine what a retiree deserves. Future retirees, then, aren’t at risk of receiving nothing, though without that discussion, it’s possible that they won’t receive much. The cynic in me now asks: can you think of a better way to destroy Social Security than that?


*Tax equivalence suggests that  when employees have less bargaining power they pay a majority of the employer tax.

**Some people make more than $118,000 per year, so this is an upward limit.

***Another, more accurate, way to think about this is that <12.4% of all value annually produced is set aside for retirees. With or without Social Security, retirees need food, shelter and the like to continue existing. The government guarantees a large part of these necessities by extending property rights to retirees in the form of a $1180 monthly Social Security check. In the absence of Social Security, retirees might save more money throughout their working career, but this has no effect on 12.4% of all value created being directed to people who no longer create value. After all, my spending = your income.

Calling The Bailouts Profitable Is Dishonest

Politico published a piece this week by ex-Geithner ghostwriter Michael Grunwald extolling the virtues of large banks and dismissing legislation that seeks to end “Too Big To Fail” (TBTF) as crackpot hysterics. If you’re going to argue in favor of TBTF, then you have come up with reasons why we should be grateful for JPMorgan and Bank of America. To that end, kudos to Grunwald for not backing down. We have mega-banks to thank for swallowing the Bear Stearns and Washington Mutual’s of the world, for financing global mega-projects and mega-mergers (yay!), and for giving reckless traders the opportunity to lose $6 billion in a day. The problem, Grunwald says, is not size, but leverage, and on that end Dodd-Frank and Basell III’s more stringent capital requirements are working. Interestingly, Grunwald’s proof is a link to another article he authored. It summarizes a Government Accountability Office (GAO) study that shows waning expectations of government support, though it has nothing to say about capital requirements. Perhaps I wasn’t supposed to click on it.

At a later date I might offer an exhaustive critique of Grunwald’s TBTF defense. For now, I’d like to address a oft-mentioned claim that sneaks into the middle of the piece, namely that the financial bailouts turned a profit for taxpayers. Here, like before, Grunwald links to a self-authored piece from 2010 that doesn’t really defend the profitability point he’s trying to make. At the time of his 2010 writing, it was being widely reported that the federal government had returned $25 billion in profit on the first $300 billion that banks had paid back. That made for a great talking point for bailout apologists. However, as some commentators made note of, the biggest loans hadn’t yet been repaid, and it was possible that when all was said and done the federal government would lose money. It just so happens that all is now said and done, and while the bailed out banks have returned more nominal dollars then were lent to them, the supposed bailout profit dwindled to $15 billion.

Despite his shady use of hyperlinks, Grunwald is likely aware of the $15 billion figure. That sounds like a lot of money until you compare it to the $426 billion the federal government injected into bailed out companies. Compared to the size of the bailout, the federal government over six years received about a 3.5% cumulative return. That means their investment yielded a 0.57% return every year. Moreover, those figures are nominal, so the real return is even smaller. I suppose you could still stick to the $15 billion figure because $441 billion is indeed more money than $426 billion, but calling it profitable is a stretch. Would anyone consider it profitable to lend someone $10 today, only to be paid back $10.35 in 2021? Of course not. The bailouts can still be defended if it runs at a loss to the federal government – after all the government’s responsibilities extend well beyond making money – but it’s lazy and irresponsible to pad those arguments with eye-grabbing claims about how it turned a profit for “taxpayers”.

Taking Supply Chains Seriously

Now that I have gainful employment I’ve transitioned from renting books at the library to purchasing used ones online. Internet retailers are cool for nerds because they allow you to track in real-time the geographic movement of your package. The last book I bought, for example, started somewhere in the Northwest, was shipped to a DHL eCommerce facility in Auburn, WA, was then loaded onto a plane to be flown across the country to another DHL facility in Franklin, MA, after which DHL tendered it to the local post office in Franklin, who then drove it about 40 miles northeast to a post office in Brookline, MA where it was finally sorted and delivered to my doorstep. The whole process took about six days. Judging by the intricate journey traveled it probably involved the coordination of hundreds of workers. Did I mention that I paid $3 for the book with free shipping?

There’s nothing unique about the delivery route described above. Millions of packages make similar trips everyday. Many of the most profitable businesses in the world rely on complicated logistical networks to produce and deliver goods that workers produce. In industry parlance the combination of production centers, shipping routes, insurance contracts and finance capital necessary to sustain the system is called a supply chain. Supply chains are the pulse of the global economy. As such, they occupy a strategic battlefield for organizers intent on wrestling power away from capitalists and into the hands of workers.

A sophisticated global supply chain’s impact is greater than the convenience it offers wealthy consumers or the profits it yields business; it also introduces several degrees of separation between the purchasers of consumer goods and the workers who produce them. In a wonderful feature – the sort of stuff viewers can reliably expect from him – John Oliver explores the many ridiculous ways this separation manifests itself in the clothing industry:

TL;DR – The cheap clothes Americans love to buy and discard frequently are the product of especially exploitative labor conditions in countries like Bangladesh. Every once in a while, something so ghastly happens that it captures the attention of Western buyers. This leads to pressures on corporations selling these clothes (GAP, H&M, Zara, etc.) to introduce less inhumane standards in the factories they source from. To that end, they have failed utterly and completely.

Global supply chains give business a veil to hide behind. Wal-Mart, for example, doesn’t own or even manage the factories in poor countries that manufacture Wal-Mart products. Instead, their warehousing and production needs are contracted out to companies who often subcontract work to other companies, and so on and so forth. What Wal-Mart does do is use their market power to squeeze subcontractor’s margins, and in so doing encourages lower wages and more dangerous workplaces. Consumers get cheap goods, laborers in poor countries get long hours in factories without fire extinguishers, quid pro quo.

What can people not complacent with sweatshop labor subsidizing their well-being do to alter this balance? What might a strategy that makes corporate behemoths more vulnerable look like? Research suggests that power resides in the subcontracting relationship. In her 2010 book Labor Rights and Multinational Production, political scientist Layna Mosely argues that “the ‘mode of entry’ employed by multinational corporations (MNCs) has important – and under-theorized – consequences for workers’ rights and working conditions”. Her argument extends the distinction between “climbs to the top” – where foreign direct investment (FDI) raises productivity and wages – and “races to the bottom” – where trade competition puts downward pressure on labor rights – by suggesting that how MNCs structure their production networks in poor countries impacts how laborers collective rights. Her empirical work shows how countries that receive production from MNCs that is offshored (via FDI, owned by MNC) have higher labor standards than countries integrated in the global economy through outsourcing (subcontracted facilities, not owned by MNC).

The intuitive explanation for Mosely’s findings is that MNCs are more accountable for working conditions when they own the facilities manufacturing their products. Subcontracting, on the other hand, allows MNCs to pass the buck to small factory owners in other countries. Even if GAP tries their hardest to clamp down on abuse in Bengali factories producing GAP jeans, they can only do so much. Or so they say.

The offshoring/outsourcing division has become extremely relevant for labor advocates in this advanced stage of global capitalism. Fast food workers in the United States are facing an analogous situation in the form of franchising. Battles over ownership structure determine the field labor struggles are fought on. It’s high time for the Left to take that battle seriously.

What Does It Mean To Be A Welfare Queen?

Ask my father’s generation what it means to be a ‘welfare queen’, and perhaps they will reference the infamous Linda Taylor, who reportedly “used 80 names, 30 addresses, [and] 15 telephone numbers to collect food stamps, Social Security, veterans’ benefits for four nonexistent deceased veteran husbands, as well as welfare.” The subset of America being referred to is clear. Welfare queens are poor, black women with a heightened proclivity to trade pregnancy for welfare benefits. They are parasitical, producing nothing of value while taking gratuitously from the public coffer that the hardworking fill. Irrespective of job market participation and by the virtue of occupying a particular space in the distribution of ownership and goods, they are lavished with material things. The welfare queen is a textbook rent-seeker.

The conservative strategy of invoking variants of the welfare queen to drum up political support among whites is well-documented. Perhaps hopeful that white support rested more on tax fairness than on racial animosity, the left has appropriated ‘welfare queen’ for its own purposes. Instead of targeting poor, black women, the left increasingly uses ‘welfare queen’ to indict very large businesses that pay very low wages. Because welfare programs are designed to ensure some basic level of welfare, government covers the difference between wages and that basic level. Wal-Mart employees, for example, rely on $2.66 billion in government assistance every year. Absent these programs, the Walton family would have to pay laborers a wage sufficient for reproducing their labor. Thus in effect, welfare programs allow business to lower their labor costs without any corresponding fall in labor supply. Just like the (imagined) welfare queens of yesteryear, Wal-Mart captures an effective rent.

It may seem counterintuitive that subsidies granted to workers can be captured by non-workers, but ‘tax incidence’ (as the phenomenon is referred to in the academy) is an established economic reality. Underlying the theory is that more important than who a tax or subsidy is levied on is how sensitive quantities supplied and quantities demanded are to price changes. If you absolutely need to eat exactly one apple per day to continue living, then the quantity of apples you consume will not change when the price of apples increases tenfold. Similarly, if government levies a tax on apple producers to the effect of $1 per apple, apple producers will raise their price by $1 and consumers will shoulder the entire tax burden. This is true whether the tax is on producers or consumers. It’s a hyperbolic example, and it doesn’t consider income, but you get the picture.

We can understand the ‘welfare queen’ argument by applying tax incidence to the labor market. Below is a generalized graph of the labor market and how it responds to the Earned Income Tax Credit* (EITC), where S(w) shows labor supply pre-EITC and S'(w(1+e)) shows labor supply with EITC accounted for (EITC benefit = e = (w + e) – (w’)).


The graph shows us two things. First, (w+e) > (w), meaning the EITC increases the wages of workers it directly affects. Second, (w) > (w’), meaning the EITC lowers the wage businesses pay affected workers. Whereas previously business paid the ‘initial wage (w)‘, they now pay only the ‘market wage with employee subsidy (w’)‘. Workers received a subsidy, and so does business.

Thus, it appears there is some truth to the left’s appropriation of ‘welfare queen’. For equivalence, it should be noted that low-wage workers, to whom the EITC applies, also receive a benefit, lest we turn public opinion against programs that reduce poverty.

*I use the EITC because its a direct wage supplement. Programs like food stamps probably allow for big business to pay workers less, but those programs affect the shape of the labor supply curve (more elastic) rather than shifting the curve.                                                               **I should preface this entire post. Tax incidence assumes a baseline that doesn’t exist. Government doesn’t levy the EITC onto the labor market because the labor market doesn’t exist without government. The EITC is part of a larger set of institutions that establish what the distribution of resources is. Economic populism – to which ‘welfare queen’ belongs – is useful, but the arguments are hardly ever intellectually coherent.