Saturday, January 16, 2021

The Fallacy of the Heap Returns

The (il)logic of Obamaeconomics has returned:

  • Each $100 of weekly unemployment bonus has hardly any negative effect (some might say "no statistically significant effect") on employment or GDP.
  • Food stamp benefits have hardly any negative effect on employment or GDP.
  • Rental assistance has hardly any negative effect on employment or GDP.
  • Obamacare subsidies have hardly any negative effect on employment or GDP.
  • Minimum wage increases have hardly any negative effect on employment or GDP.
  • A new negative income tax has hardly any negative effect on employment or GDP.
My writings rarely express disagreement with any of the bullets by themselves.  For the sake of argument, I am not disputing them here.

Where I disagree is the assertion that the combination of all of the bullets, which is Biden's pandemic relief proposal, would have hardly any negative effect on employment or GDP.  We could talk about the convexity of deadweight costs, and the fact that the new redistribution goes on top of a lot of redistribution in the baseline, but this point was well known even among the noneconomist ancients.  It is the fallacy of the heap.

Every Econ 101 teacher has to confront this fallacy.  Invariably every class has a student pushing back on the law of demand, "Raise the price of a car by a penny and that will not stop anyone from purchasing the car."  Teacher has to say, OK let the seller of the car raise the price $10,000, just one penny at a time.  Each penny increase has no effect on sales ... therefore $10K has no effect on sales.

Economists  beware of vague predicates!  Biden's smorgasbord of redistribution will noticeably reduce employment and GDP.

Wednesday, January 13, 2021

Parler Competes Horizontally with Amazon, Apple, and Google?

Parler is a microblogging platform that was cut off from the internet by its web hosting service, Amazon.  Almost as drastic, Apple and Google took steps to prevent the use of Parler on their mobile operating systems.  Unlike Twitter, Amazon, Apple and Google do not describe themselves as microblogging platforms.  At first glance, these would seem to be vertical (i.e., supply chain) relationships with no potential to harm competition.  Wrong!

For brevity, I will not go into the competitive effects of vertical relationships; the recent guidelines jointly issued by DOJ and FTC are thought provoking.  I will also not go into dynamic effects, such as the possibility that Parler's success might ultimately prove to be complementary with "right wing"/"less woke" web hosting services that are obviously competing horizontally with Amazon Web Services (AWS).

My only point here is that an important market in the modern economy is the market for personal data.  Amazon, Apple, Google, Twitter, and Facebook profit immensely in this market.  By now everyone knows that Parler's business model involves "free speech," but if you used the service (as I have) you would be equally exhorted by the platform's attitude toward personal data.  "We never share or sell data. Privacy is our #1 concern.  Your personal data is YOURS" Parler tells its members.

If too many internet users hear and buy into this rhetoric, the incumbent harvesters of personal data -- including especially Amazon, Apple, Google, Twitter, and Facebook -- will pay more (perhaps in kind) and profit less.  From this perspective, AWS' action looks like McDonald's severing the electric lines going into Subway [sandwich] locations at a time when Subway was just gaining traction.

Whether Parler's data rhetoric is correct (I openly questioned it whereas Jaron Lanier predicted that the personal data conflict would flare up particularly among the middle class) is hardly relevant to its harms to big tech.  Analogy: How many employers would like to sponsor classes on Marxism?  Marxism, which paints employers as thieves, is wrong but it nonetheless could harm an employer if it were discussed too much.

The incumbent harvesters of personal data do not want even a slight ideological disturbance of their cash cow.  That looks like horizontal competition to me.

[update: Parler filed a reply today in Parler v. AWS.  The reply noted another type of competition between Parler and Amazon, which is related to the complementarity conditions noted in the vertical merger guidelines.  Specifically, Parler asserts that (a) Twitter is client of AWS and (b) AWS did mention termination of services, even on/after Jan 6, until it learned that Trump was banned from Twitter and thereby likely headed to Parler with millions of Twitter users.]

Monday, January 11, 2021

Updates on Parler's War with Big Tech

Amazon turned off Parler’s servers last night, disconnecting Parler.com from the internet, and with it the social media accounts of almost 20 million people.

Parler CEO also reported (yesterday AM on Fox Business) that attorneys, email, and other suppliers are refusing to supply his company.  Although he said that transferring the company's software and data could be done in about 12 hours, today he wrote that "We will likely be down longer than expected" because of the time it takes to find vendors who are willing to supply the company.  Parler investor Dan Bongino said today that "Parler will be back by the end of the week."


The Jan 6 violence and lawbreaking was reportedly organized on Facebook, but Facebook continues uninterrupted.


I estimate that Parler has 20 million members (no bots), as compared to less than 60 million US accounts on Twitter.


This outrageous situation is best handled by the market, without government interference.  To the 20 million people temporarily disconnected, I say: do not be gullible enough to think that government bureaucrats or politicians represent your interests any better than the big tech employees.  Many of them have built followings on incumbent social media and do not want to work to build that again.  Any new regulations for social media will be written by allies of the incumbents.


[This is not an official Parler communication.  I have no financial interest in Parler.]

Saturday, January 2, 2021

White House Attitudes toward "Screening Agencies" (looking at you FDA)

The Food and Drug Administration (FDA) has been part of many conversations in 2020.  To the great frustration of tens of millions, it has to approve COVID tests and arguably applies the wrong (from economic and health perspectives) standards in doing so.  It is also tasked with approving COVID treatments and vaccines.  Vaccine approvals came much quicker than experts expected, although IMO not quickly enough.

With few exceptions, economists have “long been aware that the agency causes unnecessary deaths and suffering by” its “inexcusable delays in approval” (see esp. Klein and Tabarrok's collection).  Judging from 74 years of Economic Reports of the President, the White House has not traditionally given this issue much attention.  When the FDA does appear, the sentiment generally confirms that the approval delays are harmful and need reform.

Alan Greenspan's CEA was the first in 1975, when it included a paragraph about Sam Peltzman's famous study finding that the 1962 amendments were harming consumers.  A sentence of the 1977 ERP lamented further FDA bans.

Ronald Reagan's ERPs gave the issue more attention, and more bluntly.  "Screening agencies can dramatically affect the rate of innovation. A case in point is the Food and Drug Administration" it said in 1989, as part of a three-page section on the subject (pp. 218-20).  A sentence in the 1987 ERP noted how badly FDA regulation was failing cost-benefit analysis: "the average cost per life saved varies across regulations from as little as $100,000 for NHTSA's 1967 steering column protection rule to $132 million for the Food and Drug Administration's 1979 ban on diethylstilbestrol (DES) in cattle feed." (p. 183)

Both Presidents Bush included a sentence on the issue (1993 and 2001).  Consumer harms from FDA regulation was certainly top of mind for Mark McClellan, who was a member of the CEA at the beginning of the GW Bush administration until in late 2002 when he went on to head FDA.

Janet Yellen's CEA had a full chapter about regulation (Chapter 5 of 1998 ERP), especially climate change, which generally expressed the view that more federal regulation is needed.  Interestingly, five pages of that chapter express some sympathy for the view that FDA has overregulated.  It notes (p. 188) that FDA has been biased toward stopping "unsafe drugs that may cause injury or death" at the expense of "preventing sick people from getting more effective treatment."  The bias is especially questionable, it says, in the context of a life-threatening illness.

The historical context may be especially relevant to understanding ERP 1998: Yellen's boss President Clinton was dealing with a Republican Congress led by Newt Gingrich.  Both Republicans and AIDS advocates wanted to reduce FDA approval delays (the FDA review times for the AIDS treatments were still much longer than those under 2020's Operation WARP Speed), sponsoring the 1997 FDA Modernization Act that President Clinton signed.  The FDA would ultimately, as Scott Gottlieb put it, “steadily disregard[] many of the law’s provisions.”

I see only two exceptions in the 74 ERPs.  President Kennedy signed the Drug Efficacy Amendment that Peltman would later find to be so harmful.  The 1963 ERP characterizes the amendment as "protecting public health."

The second exception is the Obama Administration.  The 2012 ERP cites the FDA as a prototype of "a Smart Approach to Regulations."

By comparison to all previous administrations, President Trump's White House was arguably obsessed with FDA harms to medical innovation from the very beginning.  FDA critic Scott Gottlieb was immediately appointed to head the FDA.  The 2018 ERP had a full chapter about the health sector, half of which was about "Improving People’s Health through More Access to Medical Innovations" and "Encouraging Innovation, and Making It Affordable."  The 2019 ERP (p. 18) cites FDA deregulation as one of the highlights of the year, and devotes twelve pages to how FDA reforms increased competition and reduced prescription drug prices.  The same report also looks at the possible negative innovation effects of a proposed Federal ban on for-profit healthcare, which even Vox acknowledged as "a plausible downside."

The 2020 ERP updated the status of the FDA reforms in its chapter about deregulation, its chapter about healthcare, and its chapter about competition policy.  It also cited the new Right to Try law, and regulatory barriers to treating chronic kidney disease.  FDA barriers were part of the discussion we had with President Trump when he was signing the 2020 ERP (see the photo below).  He clearly indicated to us that his experiences with FDA so far would be helping him remove still more barriers.  (That promise was kept within weeks when Operation WARP Speed launched).  


Citing a pandemic vaccine study finished in September 2019 (sic), page 193 of the 2020 ERP estimated that "the cost of delay in vaccine availability in the case of a pandemic is $41 billion per week."  As White House senior staffer Joseph Grogan put it, the September CEA report was “was part of the intellectual foundation to the modernizing pandemic vaccine production executive order that [the White House] did along with Tony Fauci and NIH.”

[For context versus the current pandemic, note that the 2020 ERP went to the printer in early January 2020 and the writing stopped in early December 2019.  In late March 2020, knowing about the COVID-19 pandemic, I estimated that vaccine delay would cost just the U.S. as much as $136 billion per week].