The Top-Down Reformer’s Calculation Problem

Two recent examples.

1. I was invited to attend the Progressive Policy Institute on Wednesday, but not as a speaker. The topic is introduced by saying

Now that Congress has passed the Every Student Succeeds Act (ESSA), states are revamping their federally required systems to measure school quality and hold schools accountable for performance. But most are doing so using outdated assumptions, holdovers from the Industrial Era, when cookie-cutter public schools followed orders from central headquarters and students were assigned to the closest school.

In today’s world, that is no longer the norm. We are migrating toward systems made up of diverse, fairly autonomous schools of choice, some of them operated by independent organizations, as charter, contract, or innovation schools. Before revising their measurement and accountability systems, states need to rethink their assumptions.

2. And David Cutler must be happy to read this story.

Medicare on Friday unveiled a far-reaching overhaul of how it pays doctors and other clinicians. Compensation for medical professionals will start taking into account the quality of service – not just quantity.

A Nobel Prize in economics was just awarded in part for the insight that it is a bad idea to compensate workers on factors that are heavily influenced by luck. In my view, having someone in Washington evaluate a school or a teacher or a doctor does exactly that.

People who are close to the schooling process, including parents, peers, and principals, can use judgment to evaluate teachers. That’s the way it used to work 50 years ago, before the advent of consolidated, unionized school districts.

For doctors, the prevalence of third-party payments means that their compensation is being determined by remote bureaucrats regardless.

Factor-price Non-equalization?

A commenter wrote,

Given this definition of ‘discipline’, under what condition do you stop believing your intuition? What would you observe that would cause you to drop your belief in price-factor equalization, or, assortative mating?

Coincidentally, Josh Zumbrun writes,

Why would a company pay someone $80,000 if most people with an identical background—clones, in the paper’s parlance—earn $40,000? Conversely, why would someone with that background stay in the job earning $40,000 if another company will pay $80,000 for the same work?

The puzzle is that worker pay increases are highly correlated with the rate at which profit increases at the firms that they work.

My thoughts:

1. I tend to distrust the ability of economists to know more than a firm what the firm’s workers ought to be paid. Imagine an economist trying to tell Google that, based on your regression equation, it is paying its programmers more than the market wage. Google might reply that its programmers earn more because Google has selected better programmers.

2. The article offers the theory that firms with monopoly power will pay workers more. Perhaps, but a monopolist still has an incentive to avoid paying an above-market wage to its workers. In any case, if Google pays more because it has monopoly power, how pervasive is this? Do they pay above-market wages to janitorial staff? Above-market prices for office supplies? When Google employees travel, does Google pay more than the asking prices for hotel rooms and airline tickets?

3. Suppose that we grant that two workers who appear to be identical to someone running a regression equation are in fact identical in practice. It could be that each worker made a long-term commitment to a firm, and one chose a firm that happened to succeed and the other chose a firm that happened to fail. That might lead to factor-price non-equalization.

As you can see, I am very reluctant to let go of factor-price equalization. But if more evidence against it accumulates, I will be willing to change my mind.

However, in Specialization and Trade, I make the point at length that economic propositions are not falsifiable. If you want to stay committed to a proposition, you can. I say that economics deals with very few falsifiable hypotheses and many non-falsifiable frameworks of interpretation.

I can tell from the comments on previous posts that my position bothers some people enormously. They strongly oppose the situation as I describe it. Perhaps it will help to say that I have nothing against rigorous falsifiability in science, I just do not think it can be carried out in economics.

I hope you can appreciate that this is the situation with history. I doubt that we will ever be able to prove that wars are caused by X or that revolutions are caused by Y. Still, there are useful interpretive frameworks that tell us something about these issues.

I believe that all of the disciplines that deal with human society are going to have to live with this. If someone is really attached to their interpretive framework, it will be difficult or impossible to dissuade them. The best that one can hope is that people will not be so unreasonable as to assign 100 percent credibility to any information that supports their view and 0 percent credibility to any information that opposes it.

I think that we naturally try to fight back when one of our views is threatened, as mine are by the research cited above. The ideal would be for us to fight back when a study supports our views and be more accepting of studies that threaten our view, but it is difficult to live up to that ideal.

Why Do Firms Exist?

Kevin Bryan writes,

A perfect theory of the firm would need to be able to explain why firms are the size they are, why they own what they do, why they are organized as they are, why they persist over time, and why interfirm incentives look the way they do. It almost certainly would need its mechanisms to work if we assumed all agents were highly, or perfectly, rational. Since patterns of asset ownership are fundamental, it needs to go well beyond the type of hand-waving that makes up many “resource” type theories. (Firms exist because they create a corporate culture! Firms exist because some firms just are better at doing X and can’t be replicated! These are outcomes, not explanations.) I believe that there are reasons why the costs of maintaining relationships – transaction costs – endogenously differ within and outside firms, and that Hart is correct is focusing our attention on how asset ownership and decision making authority affects incentives to invest, but these theories even in their most endogenous form cannot do everything we wanted a theory of the firm to accomplish. I think that somehow reputation – and hence relational contracts – must play a fundamental role, and that the nexus of conflicting incentives among agents within an organization, as described by Holmstrom, must as well. But we still lack the precise insight to clear up this muddle, and give us a straightforward explanation for why we seem to need “little Communist bureaucracies” to assist our otherwise decentralized and almost magical market system.

Read the whole post. Pointer from Tyler Cowen.

I still think that Alchian-Demsetz is the best place to start. Suppose that a bunch of computer programmers, loan officers, and bank tellers get together to start a bank. They cannot just bargain with one another on roles, responsibilities, and pay. You need a decision-maker. And that decision maker must serve a definitive owner. The owner is the “residual claimant” on the firm.

I think that this is similar to why we certain key components of infrastructure are centralized. Road systems, sanitation systems, communication wiring, and the electric grid, for example. Imagine a bunch of households get together and say, “Let’s have a road system.” They cannot just each decide to build roads in the vicinity of their homes and then bargain with one another on roles, responsibilities, and tolls to charge drivers. You need a decision-maker. etc.

I think that we know intuitively why firms exist. The challenge is to articulate that intuition.

Hart and Holmstrom win the Nobel

1. Alex Tabarrok and Tyler Cowen (also here) have the most useful posts.

2. I was very relieved that the prize did not go to a macroeconomist. I do not see how I could write a charitable post if there is another Nobel Prize given for macro. I especially did not want to have to write a post about Bill Buckner getting into Hall of Fame.

3. On Holmstrom, Alex writes,

Suppose that you are a principal monitoring an agent who produces output. The output depends on the agent’s effort but also on noise. It wouldn’t be a very efficient contract to just reward the agent based on output since then you would mostly be responding to noise–punishing hard-working agents when the noise factors were bad and rewarding lazy agents when the noise factors were good. Not only is that unfair–if you setup a contract like this the agents will a) demand that you pay them a lot of money in the good state because they will be taking on a lot of risk that they don’t control and b) the agents won’t put in much effort anyway since their effort will tend to be overwhelmed by the noise, either good or bad. Thus, rewarding output alone gets you the worst of all worlds, you have to pay a lot and you don’t get much effort.

As I read Cosmides and Tooby, hunter-gatherers understood this. You want to let gatherers keep their output, which is not luck-driven, but you want hunters to share output, which is very dependent on luck.

Sometimes, theorists use a lot of math to come up with results that people operating in the real world have arrived at through experience. Indeed, this is a good thing, in my view. Because often the alternative is to come up with results that have no real-world relevance at all.

4. Maybe Cosmides and Tooby should get some consideration for the Nobel.

Self-driving Cars and Car Ownership

Joshua Gans writes,

It seems that the sharing economy was a transitional state from private ownership to corporate ownership. The point is that if the technology that allows sharing is good enough, the incentive to own a car — even to rent it out — goes down. And it goes down potentially all the way to zero.

Here is how I look at it:

1. What you want is car rides on demand. A few years ago, you needed to own a car in order to get that, because taxis can take a long time to show up and they are expensive.

2. If Uber and Lyft are reliable and affordable, then you do not need to own a car. The drivers own the cars. But in theory Uber and Lyft could own the cars–if they want to put up with the hassle of figuring out how to make sure drivers do not abuse the vehicles.

3. If autonomous vehicles are possible, then Uber and Lyft do not have to worry about what the driver might do to abuse the car, so it makes sense for the corporation to own the cars.

Price Discrimination Explains Everything

A commenter writes,

Standard economics that is taught in intro economics says that higher prices induces greater supply. Under certain restrictive assumptions that is generally true.

But there are exceptions. One is when the bulk of the firms costs are fixed and variable costs are a minor cost of bring the product to market

I tend to think of this as the rule rather than the exception, and I agree that introductory economics classes should give it more attention. I would say to my high school econ students that price discrimination explains everything. By that, I mean that in real-world business, the challenge is often recovering fixed cost. Marginal cost is often low, so you want to charge a low price for use on the margin. But you want to try to extract a high price from people willing to pay.

For one example among money, if you are a cell phone service provider, rather than charge the same price per unit of data for all customers, you try pricing tiers. $X per month for 0-2 gigs, $Y a month for 2-4 gigs, etc. You are trying to recover more of your fixed costs from the big users, and at the same time your users feel that the marginal cost is zero, at least until they bump up against the next tier of usage.

Cable company bundling is another classic example of price discrimination explains everything. So are store coupons. So is the price of popcorn at movies.

Evidenced-Based Policy

Robert Doar writes,

Momentum for evidence-based policymaking is building at all levels of government, from federal legislation funding rigorous evaluations to the bipartisan Commission on Evidence-Based Policymaking to counties looking to make funding decisions based on results.

I am afraid that my reaction is to be cynical. When you make funding decisions for programs based on evidence, what will change will be the reported evidence, not the programs.

During the Vietnam War, Secretary of Defense Robert McNamara was famous for demanding statistical evidence that strategies were working. He got what he was asking for, but the statistical evidence did not capture what was really happening.

To cite another example, the Stiglitz-Orszag paper on Fannie Mae and Freddie Mac appeared to be evidence-based. Recall that they wrote,

This analysis shows that, based on historical data, the probability of a shock as severe as embodied in the risk-based capital standard is substantially less than one in 500,000 – and may be smaller than one in three million. Given the low probability of the stress test shock occurring, and assuming that Fannie Mae and Freddie Mac hold sufficient capital to withstand that shock, the exposure of the government to the risk that the GSEs will become insolvent appears quite low.

Within any organization, including a profit-seeking business, one has to be cynical about “evidence.” Show me a CEO who always believes every report he or she receives from middle management, and I will show you a company that is at high risk for going bankrupt very soon. I have never been a large-company CEO, but if I were I would make a point of setting up internal checks and balances so that I did not have to rely on any one set of carefully crafted reports.

You are entitled to ask, “How can you be against evidence? Evidence is bound to make policies better than if evidence is ignored.”

My response is that I am afraid that evidence will be distorted to make spending programs and regulations appear better than they really are. I will take public choice theory over misleading evidence, any day.

Two Random Thoughts on Artificial Intelligence

I attended a talk and two panels, one of which was moderated by Alex Tabarrok.

1. Susan Athey said that companies like Facebook and Google are learning rapidly by doing many large randomized controlled trials. This gives them a way to lever their leadership positions. It suggests that “deep learning” might boost economies of scale.

2. Colin Allen suggested that if self-driving cars are programmed to stop for pedestrians, and pedestrians know this, pedestrians could become more reckless and aggressive. Hmmm.

My Early Prediction of I-Phone’s Latest Innovation

When I saw this story,

Now it’s headsets: spending on wireless headsets overtook wired ones last year, says Steven LeBoeuf, founder of Valencell, a developer of biometric sensor technology for wearable devices.

I was reminded that back in 2001 I wrote,

I can imagine a world in which everyone spends several hours a day wearing a headset. There will be a software industry devoted to building applications for the headset platform, which consists of earphones, a microphone, and something that I call a “tuner.”

It is an amusing essay to read–wrong in many respects, but actually quite prescient in others.

Timothy Taylor on the Tech Sector

He writes,

My own guess is that the applications for IT in the US economy will continue to be on the rise, probably in a dramatic fashion, and that many of those applications will turn out to be even more important for society than Twitter or Pokémon Go. The biggest gains in jobs won’t be the computer science researchers, but instead will be the people installing, applying, updating, and using IT in a enormously wide range of contexts. If your talents and inclinations lead this way, it remains a good area to work on picking up some additional skills.

This sounds right to me. The technological advances have been rapid, but the process of deploying applications goes more slowly.