12 May 2016

The Case For Panic

Would A Trump Presidency Be Apocalyptic?

Paul Campos, at the Lawyers, Guns and Money blog, makes the case that a Trump Presidency would be such an Apocalyptic event that we should worry about it, even if it isn't very likely.

I'm not willing to go so far as the A-word when predicting the likely consequences of a Trump Presidency.

Simply put, too many barriers stand in the way of major Trump driven screw ups in our political economy that buffer us from disaster at the hands of a bad President, particularly when he doesn't have widespread support in the rest of the political system.  These include:

(1) A tradition of rule of law enforced by strong and independent judicial system,

(2) separation of powers between the President and Congress,

(3) an independent central bank,

(4) a moderately independent FBI,

(5) a civil service that is functionally insulated from the Presidency in many key respects,

(6) federalism that leaves lots of important governmental functions to state and local governments,

(7) an economy that has a relatively small federal public sector (many government owned enterprises are a the state and local level and the federal ones are quite independent of the President),

(8) the Posse Comitatus Act,

(9) a history of military non-involvement in civilian politics,

(10) a federal legislative process in Congress that makes it hard to change the status quo and doesn't give the President much formal power to positively enact new law, and a regulatory process under the Administrative Procedures Act in the federal government that mutes and delays Presidential influence on the federal rule making process,

(11) Trump's lack of support from leaders of most of the traditional international allies of the U.S., and

(12) Trump's lack of support among the elites of either of the major political parties, so he can't count on them cooperating with or conspiring with him to break traditional rules limiting Presidential power.

Trump's immense personal wealth, while it is a huge asset on the campaign trail, is also something that would provide only marginal benefit to Trump as a sitting President.

Also, don't forget that if Trump really screws up, referendum type swing voters, who make up a significant share of independent voters, are likely to join with Democrats to vote him out of office after just four years.

And, if Republicans in Congress appear complicit in working with Trump to ruin the country and enough voters interpret the situation that way, voters could replace Republicans with Democrats in mid-term elections two years after Trump takes office (to some extent this is precisely what happened when Republicans won the 2010 mid-term Congressional elections in the wake of dissatisfaction with Obamacare in many swing Congressional districts).

While somewhat outdated, consider this analysis from an October 25, 2005 post at this blog:
As of 2002, the last year for which I have good figures easily at hand, there were about 22.8 million governmental employees, military and civilian combined, in the United States. 
Only a small fraction of them report to the President. 
About 18.6 million of those employees worked for state and local government, (about 82% of the total). 
More than 37% of the civilians employed by the United States government also do not report to the President, because they work for Congress, for the Judiciary, for the Postal Service, or as one of the roughly 182 thousand other employees of independent agencies (this is really somewhat more complex, because some independent agencies like the EPA and NASA report pretty directly to the President, while other agencies within the Executive Branch, most notably the FBI have considerable independence from the President who cannot, for example, fire the FBI director without good cause). 
Thus, only about 11% of governmental employees in the United States report directly to the President and about 46% of those are military personnel. 
Only 9,051 people who work for the United States government in the executive branch are political appointees. A significant share of these appointees, moreover, are to insignificant posts like the Barry Goldwater Scholarship and Excellence in Education Foundation, or the United States Holocaust Memorial Council. There are, for example, only eight political appointees in the Central Intelligence Agency, widely believed to have about 30,000 employees. 
Everybody else in the government is either a merit system employee or hired by someone who is a merit system employee. No bid contracts are the exception, rather than the norm.
A President facing an uncooperative U.S. Senate can also have difficulty getting his political appointees approved and on the job.

Trump's Odds Of Winning May Be Underestimated

Campos further argues than the conventional estimates of the likelihood of Trump winning in the general election (for example based upon state by state general election contest polling) may very well be underestimates, particularly in light of the fact that Trump seems to do better in polls that are anonymous and computerized as opposed to those were a human interviewer is involved.

Campos starts with a 10% seat of his pants expectation of the odds of Trump winning and then goings onto identify factors that would elevate that probability without quantifying them.

Political science professors Steve Greene, the CNN affiliated prediction market, and I have all put Trump's odds of winning in the 20something percentage range, and the Campos post, while not putting a number on the odds seems to end up there as well.

Certainly, almost nobody thinks that Trump is more likely than Clinton to win the general election at this point, but the odds of that happening aren't that remote.  You wouldn't cross the street if your odds of being hit while doing so were anywhere near as great as the odds of Trump becoming our next President.


When the odds of a horrible President are somewhere between one in five and one in three, there is indeed cause for concern, but perhaps not as much of a cause for panic as Campos would imply.

No comments: