You Need Rules

This piece about Google Guy from Michael Brendan Dougherty was filed a week ago, but I only now have enough time to respond to it. In the post, Dougherty vents his frustration with the argument that conservative labor market institutions are responsible for the termination of Google Guy.

The argument he criticizes goes like this: Google was only able to fire the guy because our employment laws give Google discretion to fire anyone for any reason or no reason at all. If we had a different set of laws that only permitted Google to fire someone for incompetence or economic redundancy, this could not have happened.

Dougherty’s response is to just say that, though it is true the laws give Google the power to make this kind of move, Google only made the move because of the pressure it received from liberals. If liberals had not pressured Google, then it would not have exercised its power to dismiss him.

Dougherty sees this as a rebuttal to the argument about labor market institutions, but it is perfectly compatible with that argument. Once Dougherty’s insight is added, the synthesized argument becomes: Google only fired the guy because 1) it had the legal power to do so and 2) it received pressure to do so. If either (1) or (2) is removed, the termination does not occur.

Given this understanding of the situation, what exactly are we supposed to conclude? If you think, as Dougherty does, that the termination was unjust, what would he actually have us do about it?

Presumably, he would want a state of affairs where Google retains the power to arbitrarily dismiss anyone but people just decide not to pressure them to use that power. Though this kind of posture makes great fodder for a blog post, it is completely ridiculous in reality.

For as long as Google has the power to do something others want them to do, those others will pressure Google to exercise it. This is true broadly. Whenever you allocate power to some person, at least some people will pressure that person to use it in the way they want to see it used. And this is not a left or right thing. It is true across the board and a lot of pressure to apply power in one way or another has nothing to do with anything remotely political.

If you really want to cut this kind of thing out, you need to replace dictator-like power with rules. Rules will strip Google of the power to act and give it an ironclad reason to refuse to act. When rules are in place, Google would turn to the pressure groups and say “sorry folks, our hands are tied.” And, once the rules are well-established, it wouldn’t even have to do that because nobody would bother to pressure them in the first place, knowing it was futile. By removing the power to fire, you remove the demands to fire, killing two birds with one stone.

I understand why libertarians and management types object to rules requiring a “just cause” for termination. But I can’t figure out why more economically squishy social conservatives like Dougherty object to them. From the outside, it seems like yet another instance of social conservatives being taken for a ride by the business wing of the conservative coalition.

The Trump NLRB Will Smash the Google Guy

The Google Guy who wrote the memo about Google’s diversity efforts has provided great fodder for the take-makers. Was he doing free speech? Was Google doing political correctness? Can a non-government actor hurt free speech? These and other profound questions about the nature of liberalism are all at stake.

Curiously enough, nearly every person’s opinion about the correct application of underlying speech principles in this case can be predicted by their substantive political orientation rather than their views on liberalism per se. The same of course was true in the Kaepernick case before this one, or the Duck Dynasty case before that, or the Dixie Chicks case before that. If you were inclined towards cynicism, you might even conclude that nobody actually cares at all about abstract procedural liberalism, that society is simply an unending power game, and that nobody is actually motivated by the arguments they write for publication.

Thankfully, we don’t even have to consider any of that here because this post is not about that. It is about the Google Guy’s chances at the NLRB, a topic I have been urged to write about by multiple people.

I said on Twitter and still believe that, under current NLRB law, the Google Guy has a good chance of being reinstated for the following reasons:

  1. The Google Guy complained about working conditions to other employees, which is generally protected activity under Section 7 of the NLRA.
  2. Although the Google Guy used corporate email to communicate with other employees, the NLRB ruled in Purple Communications that you can do that.
  3. Although the Google Guy’s comments were offensive to many, they probably are not so offensive that they lose protection of the NRLA under the NLRB’s Atlantic Steel test. For example, this statement from Pier Sixty, LLC is not so offensive as to lose protection: “Bob is such a NASTY MOTHER FUCKER don’t know how to talk to people!!!!!! Fuck his mother and his entire fucking family!!!! What a LOSER!!!! Vote YES for the UNION!!!!!!!”
  4. Although the NLRB has never directly confronted the issue of what to do where the way someone has engaged in protected activity could also be a violation under Title VII of the Civil Rights Act, the Administrative Law Judge in Arthur Young & Co. dismissed an employer’s attempt to use Title VII to defend its termination. The judge rejected the defense in a short footnote, saying the employer’s assertion that there may have been a Title VII violation was implausible. My best guess is that the claim this email represents a Title VII violation is similarly flimsy and thus a defense that said “we had to fire him because we needed to protect ourselves from a Title VII lawsuit” would be similarly rejected by the current members of the NLRB.

But this analysis will all change once Trump’s appointees to the NLRB come in:

  1. The Trump NLRB will want to overturn Purple Communications, which was decided in 2014 and reversed prior NLRB precedent on the question of the use of corporate email. The pre-2014 rule favored by conservatives says that corporate email systems are company property and so employees have no right to use them in ways not sanctioned by the boss.
  2. The Trump NLRB will also want to give employers as much latitude as possible in firing someone for offensive remarks, and so may even be willing to interpret Atlantic Steel to permit termination in this case. By lowering the bar for what counts as too offensive to be protected, they will make it easier for employers to find ways to get rid of union activists.
  3. Since the Title VII argument has not been directly confronted by the NLRB before, it is also conceivable that the Trump NLRB would rule that employers have the right to terminate employees engaged in protected activity where there is any colorable Title VII claim resulting from the way that activity was conducted.

In basically all cases, a conservative NLRB will want to reduce the ways workers can coordinate with one another, and increase employer discretion to terminate employees. When I raised this point on Twitter, someone said that this might be different under Trump because wouldn’t such a ruling feed into the political correctness and whatnot that he hates. And to that I can only laugh: at the end of the day, what conservatives want to do is shift power to bosses over workers, and they are really good at keeping their eyes on the prize.

Has Underemployment Among College Graduates Gone Up?

Steve Matthews has a piece in Bloomberg arguing that the underemployment rate among young college graduates remains elevated and that this is a sign that the labor market is unhealthy and that labor market slack still remains. I am sympathetic to the idea that the labor market remains unhealthy and that slack still remains, but this particular argument for that conclusion does not seem very compelling.

Here’s Matthews:

The relegation of college graduates to non-degree positions was once seen as a temporary blow for young people unlucky enough to graduate around the time of the deep 2007-2009 recession. Instead, millions of Americans like Reyer continue to face the same struggle.

About 44 percent of recent college grads were employed in jobs not requiring degrees in the final quarter of 2016, not far from the 2013 peak of 46 percent, while the share of that group in low-wage positions has held steady, data from the Federal Reserve Bank of New York showed Wednesday.

That’s a sign that the nation’s labor market isn’t at full health, despite an unemployment rate forecast to remain at 4.7 percent in March, close to the lowest in almost a decade. In fact, the elevated level of college grads in non-college jobs could mean there’s still slack and that the Fed can go slow in raising interest rates, betting that more high-wage jobs will materialize. It could also mark a more permanent shift in employment that the Fed can’t fix and be a tough challenge for President Donald Trump and Congress.

Along with that text, Matthews shares this graph, showing the percent of recent graduates and college graduates overall who are in jobs that do not require a college degree.


Whenever someone is making an argument about the effect of the Great Recession, I am always curious to see what the data looks like prior to the Great Recession. Starting the trend line in 2007 does not give us a good indication of what normal actually looks like.

Luckily the New York Fed runs this data series back to January 1990 on its website. The graph looks like this:

Screen Shot 2017-04-08 at 10.14.40 AM

The current figure (43.5% in December 2016) is higher than the series low point (37.6% in May 2001), but it is lower than the series high point (47.7% in December 1992). More generally, it seems as if the current figure is within the normal range.

On its face, it is kind of surprising that the figure has not budged much over the last few decades. College attainment has increased significantly over this period and so you might expect that the percentage of graduates who are underemployed would go up.

One reason why this may not have happened in this data series is the way they define what jobs require a college degree and what jobs do not.

The underemployment rate is defined as the share of graduates working in jobs that typically do not require a college degree. A job is classified as a college job if 50 percent or more of the people working in that job indicate that at least a bachelor’s degree is necessary; otherwise, the job is classified as a non-college job.

Because they define a job as requiring a college degree if the majority of people in the job say it does (i.e. it is not based on an objective assessment of the skills required), this means that jobs that used to require a high school degree but now require a college degree are scored as college-degree jobs. Put another way, this particular measure is not able to disentangle the effects of credential inflation. When the credentials for the same job go up, this measure transforms it into a job requiring a college degree and therefore a job that no longer counts as underemployment for a college graduate that fills it.