Week of December 20, 2020

Automation, Inequality, and Bank Runs

Greetings, readers, and happy holidays! The school year is over, which means that the newsletter is—for the moment—back in circulation. This has surprised me at least as it will no doubt surprise you, but I’ve been bitten by the proverbial writing bug and feel compelled to respond. I would cite the calling of a Muse, if only that did not disgrace the nature of Art by association. Instead, I invoke Butler’s earnest Ernest Pontifex:

I have been a scribbler for years. If I am to come to the fore at all it must be by writing.

Should I choose to continue with this project, even temporarily, I’ll largely be returning to the original format of links with commentary, though I may drop in an essay or two if I feel the inclination. I’ve been churning through a few papers this fall and have more in stock, and I hope that this gives me additional motivation to read more, more quickly, and more thoroughly. No promises, however, on holding to any kind of regular publication schedule (weekly output may be greater or less than one). A foolish consistency is the hobgoblin of little minds.

Without further ado, then, on to the content.

Links

Why Are There Still So Many Jobs? The History and Future of Workplace Automation

David Autor | Journal of Economic Perspectives | Summer 2015

I had been meaning to read this article for a while, mostly as a precis for the more recent Acemoglu paper on the same subject. Autor argues that, in general, automation simultaneously substitutes for and complements various tasks performed by human workers, which in turn determines the extent to which the corresponding occupations increase or decrease employment. Routine tasks such as clerical work and assembly-line production are easily automated; others, which require “tacit skills” like abstract problem-solving and interpersonal communication, stubbornly resist change. In the former case, the income effect of decreased production costs is counterbalanced by a substitution effect (of capital for labor), and whether this is a net positive or negative for employment is industry-dependent. “Tacit skill” jobs are complemented by machines, which liberate managers and professionals from mundane work and allow them to concentrate on the higher-productivity tasks in which they, as humans, have a comparative advantage. The result has been a dramatic “occupational polarization” in the US workforce, with employment rising in low-wage manual sectors (food preparation, cleaning) and high-wage professions (technicians, executives), both of which are tough to automate, and collapsing in “middle-skill” production and office jobs. Both of the poles are seeing poor wage growth, however, as a result of abundant labor supply. Autor predicts that this trend will subside, pointing to a new class of “artisanal” tasks and the persistence of interpersonal work, but the former claim remains speculative and pandemic has dampened the latter.

The Top 1 Percent in International and Historical Perspective

Facundo Alvaredo, Anthony B. Atkinson, Thomas Piketty, and Emmanuel Saez | Journal of Economic Perspectives | Summer 2013

The argument is probably familiar by now, but the authors—including two of the leading lights on inequality and taxation—have incontrovertible and increasingly disturbing data to support it. The share of income accruing to the top 1 percent of American earners has, after being halved after the Second World War, more than doubled since the end of the 1970s, from 9 percent in 1976 to 20 percent in 2011. The paper proposes several mechanisms by which this may have occurred, and intriguingly discards the technological-change hypothesis advanced by Autor (see above). The first is that declining top tax rates created an additional incentive for high-skill workers and executives to negotiate for a larger share of the “surplus” earned from business activity. This “selfish” behavior may exacerbate polarization by turning managers against firm interests and toward their own, instead of ploughing profits back into new investment. The second is through lax restrictions on inheritance flows and bequests, which permits the persistence of certain kinds of wealth especially likely to be held by high earners. Unsurprisingly, Piketty and company contend that the high marginal tax rates of the Trente Glorieuses should be reinstated—potentially justifying an optimal top tariff of 83 percent.

Modeling the Legend: The Trouble with Diamond and Dybvig

George Selgin | Alt-M | December 17, 2020

“Bank Runs, Deposit Insurance, and Liquidity” (1983) by Douglas Diamond and Philip Dybvig presents a model demonstrating how fractional reserve banking systems can succumb to runs, even under ordinary conditions. In the event of a panic, depositors slow to make withdrawals will receive sums less than the amounts to which they are entitled; thus the first sign of panic leads even financially sound account-holders to rush for the exits, fearing that their funds will be captured by the insolvent. This framework, Selgin argues, has been fallaciously applied in support of government intervention in the financial sector through deposit insurance and lender-of-last resort operations. Banks are not inherently unstable; rather, runs can be prevented by freezing withdrawals when the fraction of reserves paid out is equal to the fraction of the population with financial difficulties, assuring sound depositors that they will be repaid. Diamond and Dybvig held that this would not be possible in the event that this fraction was a random variable, recommending instead that government use deposit insurance to reward restraint and tax panickers. This, Selgin contends, is both methodologically (assuming that banks receive requests sequentially and government simultaneously) and practically unsound, encouraging risk-taking and moral hazard. Further, the paper’s main finding—that “private arrangements can't guarantee a first-best banking outcome when there is aggregate uncertainty and withdrawals are granted on a first-come-first-served basis”—is rendered unrealistic by the authors’ hypothetical bank, which ignores the potential stabilizing qualities of equity finance. Real-world institutions, whose capital is partially sourced from patient outside investors, are less likely to experience the equilibrium panic outcome to which the insurance recommendation responds. Moreover, actual depositors continually use banknotes as a medium of exchange, obviating the supposed need to convert liabilities into hard currency. Lucid and iconoclastic throughout, with many interesting links to additional research.