Thursday, July 25, 2019

On the Surprisingly Apolitical Nature of the Fed

The eternal question about the civil service is its level of competence. At one end of the spectrum is the fat, curmudgeonly woman at the DMV. At the other is the Hollywood depiction of the CIA. Reality seems to vary by department, and is usually somewhere in between. I think a lot of conservatives tend to be skeptical of government in general partly because the bits of government that they are forced to interact with are so woeful. Waiting hours in line at the DMV to fill in a form that should be able to be done online, for instance. The ridiculous and inefficient security theatre of the TSA, staffed by inept, surly, disgruntled buffoons. The post office managing to screw up deliveries at a far higher rate than FedEx or UPS. It’s only natural that this perception is extrapolated to all the bits of government that we don’t actually interact with personally.

But to a large extent, this is a function of the types of people these places hire. There’s some aspects of government that will inevitably involve distorted incentives and poor performance from a lack of competition. But even if the difference with the private sector is always there, the level doesn’t have to be appalling. In Singapore, government jobs seem to be viewed as prestigious and well-paying, and so attract relatively talented and competent people. Or, to go back further, you would give your left nut to have Evelyn Cromer administering the USA, rather than any of the leaders we’ve had since I’ve been alive. In other words, it certainly doesn’t always have to be as bad as the modern US.

When the US scrapped the civil service exams because of disparate impact (incredible, I know), it ended up having the biggest effect on low level jobs that you can’t sneak in other requirements like college degrees. This is how the DMV and the TSA got so awful – it turns out that IQ matters, even in low level clerical or customer service jobs. In this respect, the Fed has held out incredibly well by virtue of the fact that a lot of its jobs require a PhD in finance or economics from a top university. PhD programs have so far mercifully been largely spared the wrath of the Cultural Marxist need to bring in diversity even at the cost of competence. Moreover, even if you get in, you still need to pass, and convince the hiring committee that your thesis is actually good. In this sense, the Fed is largely drawing on a fairly talented pool of people who are pretty well versed in current economic research (for what that’s worth).

So if the Fed has avoided the obvious failure mode of being staffed by imbeciles, how does it fare on other measures? The interesting one is regulatory capture. Like any regulator, it can be captured by its employees, by politicians (which, ironically, is how the system is meant to work, but which in practice is usually treated as a design defect), or by the companies and groups that it’s meant to be regulating.

In terms of being captured by its own employees, this is hard to discern clearly, but I think that this has happened less than at most agencies. The biggest reason is that, other than the Fed Board, the regional Feds are notionally private, and so can set their own salaries and hiring/firing conditions. Even the Board seems to pay approximately market rates for the people it hires. This seems to gets rid of a decent amount of the insanity of the public service working conditions. When you can’t pay employees more, they extract concessions in the form of goofing off, unions to make it so they can’t get fired, etc etc. But they’d probably rather take the costs just in the form of more cash. This doesn’t have the deleterious effect that them simply being lazy has – it’s at least a transfer, rather than deadweight loss.

The biggest surprise about the Fed, however, is the fact that it seems to have been able to maintain relative political independence up to now. Independent central banks were a radical idea in the 19th century, where monetary policy was hot button political issue. William Jennings Bryan effectively wanted loose monetary policy (in the form of bi-metalism) to inflate away the debts of farmers. Letting a bunch of PhDs just run the show was probably not likely to be viewed as a compromise answer. But oddly, this kind of redistributive aspect of monetary policy doesn’t get thought about a ton anymore. Instead, the main effect seems to be about what monetary policy does to people’s 401K plans via the level of the stock market. This may be dumb short-termism, but at least everyone is on pretty much the same side.

In the modern ear, Donald Trump has decided, at least via twitter, to talk derogatively about the Fed’s policy, and suggest that they need lower interest rates. I don’t think the Fed takes this especially seriously. Which is fortunate, to be honest. Whatever you think about the Fed’s monetary policy since the great recession, you’d have to be incredibly optimistic to think that Congress or the Presidents would have done a better job. Instead, you can see exactly what the pressure would have been – lower interest rates before an election, consequences be damned. If it creates inflation, well too bad for the next guy. In other words, we could have the same level of far-sighted statesmanship that we currently observe with the US fiscal deficit, but with monetary policy as well. What a delight that would be.

At least on the monetary side, part of the reason the Fed seems to have stayed largely professional and apolitical is that it really has only one main button it can press – interest rates up, or interest rates down. And while people debate furiously over the relationship between that and the state of the economy, most people are agreed on at least the outcome they’re aiming it, namely high growth, low unemployment, and price stability, currently taken to mean low but positive inflation. It seems likely that the Fed has only an approximate idea of the relationship between the variables in question. But then again, it doesn’t seem like most of the public has any better idea either, and so are largely content to let them do what they think is best as long as things aren’t collapsing.

The main people with strong views on the matter seem to be people that want the Fed abolished and US dollars replaced with gold or bitcoin. I tend to think monetary policy when implemented sensibly is a useful tool, and giving it up for a fixed money supply would probably cause more harm than good. That said, my priors are pretty wide on what a fixed money supply would actually do for an economy. I found the Friedman case pretty convincing that letting the banks fail in the 1930s was one of the worst things the government did, and contributed significantly to prolonging the depression. But even if you disagree on this (and plenty of smart Austrians do) the Austrians’ view seems especially far-fetched on a political economy basis. When the world is melting down, governments are always going to do something, even if that something turns out to be significantly counterproductive. Even from an Austrian perspective, lowering interest rates is probably among the less harmful knee-jerk policies one could imagine, compared with, say, nationalizing industry or applying across-the-board price controls.

The more interesting question, and the one that’s harder to answer, is whether the Fed has been captured by the banks. In terms of the broad question of monetary policy, and whether and how to intervene during financial crises, there’s probably not a lot of disagreement between major banks and the Fed. If you think that they’re both wrong, this understandably looks like collusion and regulatory capture. But I think it’s more likely that both groups tended to come from the same business schools and economics departments, and this is largely what gets taught there. And while there is reasonable agreement between banks and the Fed on what should happen ex-post in a crisis (grumblingly bail out failed banks), the ex-ante question is not nearly so clear. In particular, most banks would probably like to see capital requirements cut significantly, and scrap the various costly stress tests that the Fed does on major banks. I’m not saying this is a major bone of contention, but it’s not exactly like banks get everything they want either. 

The stronger case, however, seems to involve some of the current implicit subsidies given to banks. I’m not even talking about deposit insurance, which is related to the “letting the banks fail” question above. Rather, the decision since the crisis to start paying interest on reserves looks a lot like a back-door bailout and subsidy. No no, they say, it’s just an important aspect of unconventional monetary policy. Great! So can I, as an individual deposit my own money at the Fed to take advantage of this same policy? Ha ha, no, of course not! Also, we’ll continually shut down any bank that tries to just operate as a pass-through entity to enable this, a proposal called narrow banking. When even John Cochrane is saying this makes you as the Fed look dumb and crooked, you should probably take heed.

But that’s the messy nature of regulation. You’re never going to get all of what you want, and sometimes dumb things happen anyway, usually for a mix of motives. In other words, the Fed isn’t the Hollywood version of the CIA, but it’s a hell of a lot better than it could be. I almost keep expecting it to get gutted and politicized at some point, and end up as some social justice economic group like the CFPB. It could be worse. It probably will be worse.

Contra Chinese folk wisdom, may you continue to live in uninteresting economic times.

Friday, July 19, 2019

On the Divided Nature of the Civil Service

At some point a few years ago, the predictive power of my models of permanent government improved significantly when I stopped conceiving of the government as a single monolithic entity with a single set of preferences. Rather, one usually gets a lot further by conceiving of different parts of the government as separate fiefdoms, with different aims, different power structures, and different allies.

The standard version of this is to distinguish between the politicians and the permanent civil service. It's definitely true that these groups often diverge a lot. You saw this most flagrantly during the 2013 government shutdown, where in response to mild threatened funding cuts, the Parks service put up barricades to shut down the Washington Mall, even though a) it's just some statues and grass, b) it's not clear what the Parks service even does there, and c) it takes less work to just not show up than to show up with barricades to ruin tourists' vacations. If you don't have this in your mental model of the world, I honestly don't know how to help you. But it seems that, even when just considering the actions of the bureaucracy as a whole, similar things apply. A lot of the time, there isn't even just one civil service.

For the US in particular, one occasionally runs into situations where if one is forced to ask "what does the US government want to happen?", the only conceivable description is that the US government is insane, evil and schizophrenic. To me, the cleanest example of this was the late Obama era policy towards Syria, where the US as a whole was supporting three out of the four sides of a civil war. The State Department and the CIA seemed to be gung-ho about regime change at any cost, supporting the "moderate" Sunni rebels, who kept on defecting to ISIS. ISIS was clearly undesirable, but apparently still viewed as better than Assad. While there wasn't any explicit support for them, the constant drumbeat that Assad had to go, and the fact that ISIS were almost certainly the most poised to take over the place, came across as at least tacit support. The Department of Defense was supporting the Kurdish rebels, who were fighting against ISIS. But they were also supporting the Turks, who were periodically fighting against the Kurds, because they were long-term important Defense allies.

Now, if one were forced to explain all this in terms of what "the US government" wanted, the only plausible models would seem to be that the US was just an agent of chaos, supporting everyone fighting everyone else, or that the US had gone completely mad and no longer understood the predictable consequences of its actions. If you break things down into component departments, it looks slightly less insane. Still somewhat insane, mind you, but not completely self defeating given any coherent set of policy aims.

Not only that, but you even get similar dynamics operating at times within government departments. You see this with immigration policy in Australia, and from what I hear in Canada too. Australian legal immigration is a hard-assed, skills-based, points-based system that aims to let in only people likely to be useful contributors to the Australian economy. But meanwhile, the Australian refugee processing apparatus seemed determine to wave in almost anyone, accepting 9 out of 10 claims of boat people. Including, hilariously, the captain of the people smuggling ship Captain Emad, who was also granted asylum, which suggests that they either had no ability to discern true claims from false claims, or were so worried about false negatives that they just accept any old story. If you take this seriously, it suggests that there isn't even one coherent immigration policy. Otherwise, why would it make sense to take in the highest performing migrants and every freebooter and scammer who turns up in a boat with a sob story, but not the people in between? I mean, it's not impossible to rationalise. But doesn't it just seem much simpler to posit that the refugee processing section is stuffed with bleeding heart leftists who stamp in anybody, and the other sections are much more skeptical?

This alternative theory also explains something about the Australian government's offshore processing of refugee claims. The government was eager to process claims almost anywhere except the Australian mainland - Christmas Island, Manus Island, Nauru, Papua New Guinea, you name it. Most people naively focused on the symbolism aspect - asylum seekers don't get to come to Australia! Fine, sure. But what else do all these places have in common? They're basically out of the reach of the Australian court system and existing refugee processing apparatus. It seemed to be easier for the Liberal government to set up an entirely separate parallel system of refugee processing than to reform the existing system which it was nominally in charge of.

You see a similar dynamic at play in the US as well. The only government agency that seems to genuinely support Trump is ICE, since they stand to get more money and power if his policies get through. Meanwhile, an antifa goon actually tried to assault an ICE facility with a rifle and "incendiary devices", and got killed for his troubles. The left seems to have figured out that ICE is in fact aligned with Trump, and is targeting them specifically. You sure don't see them trying to bomb the INS, which assigns citizenship to migrants.

This may sound like an argument that everything is inscrutable and you need to know endless detail about every part of the government, but this isn't really the case. Most the bits of the permanent government are at least center left in their aims and ideology, if not outright leftist. So there usually isn't that much conflict about broad objectives. Things tend to get interesting when the there's departments whose functions involve generally right wing tasks - border enforcement, the armed forces, policing, prisons. This is where the tension between the left wing slant of public servants and the right wing nature of people drawn to the particular job tend to be at odds. It's perhaps not surprising that the authors and commenters on Second City Cop seem far more sensible than Chicago's actual elected officials. They also seem substantially more sensible than Chicago's actual police chiefs of the past, who are basically political appointees.

It's not just that nobody is in charge of Moloch. It's that Moloch isn't even a single entity in charge of itself.