Institutions fail: What about it?
Over at Anti-Dismal, Paul Walker has replied to a comment I posted with a couple of posts. Now I thought I should repay the favour by writing something up 🙂
Update: Paul replies here. I agree with absolutely everything he says, but ultimately I think our views on what constitutes “optimal” action still differ as a result of different value judgments.
You should definitely go read the posts, they are very good posts which cover the idea of government failure, and the limits to market failure through asymmetric information. I completely agree with all the objective parts of his posts. However, I still believe my response to “Now there may be something wrong with the price system, but there is a lot more wrong with the government system”:
That’s really your key value judgment isn’t it. I’m not sure I agree. I am not a fan of “multipliers” – but in the face of a large, sustained, market failure I find it hard to conclude that there is no role for government.
Here’s why:
It is still a value judgment
Yes, government failure happens – a lot. Governments, like all other institutions are filled with self-interested individuals, whose actions are not fully related to the goal of the organisation.
However, in reality market failures are also endemic. There are a multitude of situations where the interests of individuals and social interests are not aligned.
Determining that government failures will always be larger than the market failures they are supposed to cover is a value judgment.
Why is your value judgment different
Market failure stems intrinsically from one thing – the failure of prices to represent scarcity. We know that markets can fail in any of the cases where perfect competition fails namely:
- Sellers of a product have market power,
- Buyers and sellers without a voice in the market are adversely/positively influenced by the market,
- Entry and exit from the market are difficult,
- Information is asymmetric,
- Institutional conflict pushes firms/buyers away from rational behaviour.
Now as Paul Walker rightly says, even when we have market failure (which we often do) we also require government action to improve outcomes – which it often doesn’t. In fact, in a number of these cases the market will adjust to improve outcomes.
However, as Paul Walker does, I would like to move away from this relatively static/short term conception of market failure and towards a dynamic view. Look again at what my comment said on the post:
in the face of a large, sustained, market failure I find it hard to conclude that there is no role for government
In Paul Walkers discussion, the price allows us to solve market failures over time. However, this is not always the case when there is interaction between agents. Specifically, there may be “multiple pareto-ranked” equilibrium which are stable and consistent over time – each with their own set of prices and welfare levels. In this case there are multiple ways the economy can go – and we have no assurance that, over time, the market price will converge to the “best” outcome.
Now, in the case of a massive negative shock to output such as the one we MIGHT have now, we are able to make a case for further government intervention to push us towards a pareto superior outcome.
For government to do this, it needs to understand the true “capacity” of the economy to produce and then work to influence “expectations” to get us to a better outcome.
Given the relationship between labour and other inputs, I believe that an unemployment rate in double digits (or even high single figures) indicates that we are at risk of moving to a pareto-inferior equilibrium. In this case, government actions seem justifiable – even if their implementation will be wasteful.
However, of course this is a value judgment on my part. Given that we will never face the counterfactual of our actions we can’t even test these value judgments – so I do not expect to convince anyone of my point of view. Even so, this should make it clear that the idea that government failure always exceeds market failure is a value judgment – it is an opinion based on belief rather than an objective statement of how the world works.
Note: My opinions on a stimulus here seem to support some degree of stimulus in the US. But unemployment in NZ is only 4.2%, even the worst forecast pick it to reach 7%. As a result, there isn’t as large a case for a stimulus package here. Furthermore, I did not cover the “optimal type” of stimulus. Ultimately, multiple equilibrium come from identified structural issues, and specific market failures – a “stimulus” package should target those, instead of arbitrarily throwing around and wasting national resources.
Matt. Thanks for your response. I have written a reply here.
Hi Paul,
Thanks for your response, it sounds good – I think we pretty much agree, but it is undoubtedly useful for us to flesh out the essential issues. I’ll try to give it a closer reading after work tonight – this whole economic crisis has made me quite busy 😉
Matt I’m not persuaded by your position for a couple of reasons. First, while the present crisis as a market failure is large I am not convinced it will be sustained in the same way that other alleged market failures (lemons, keyboards) can persist indefinitely, at least in theory. Doesn’t the “large, sustained” criterion you define here rule out a government response to the present crisis? If the current crisis does count – then what kind of market failure is ruled out?
Second, I’m not sure why you raise multiple equilibrium. Unless I’m mixing up the order of responses and replies, Paul has pointed out the evidence that economies seem to consistently converge on efficient solutions when multiple equilibria exist, and that it is hard to find actual examples of persistent market failures. People find it hard to resist the opportunities missing markets create. In fact I think that defines entrepreneurship. It would be helpful if you could epxlain why you think the market responses to asymmetric information elsewhere might not be expected to arrive now. There is, afterall, a lot of money to be made by whoever figures out the real value of US banks and real estate.
Third, for what its worth, I’m quite sure neither yours nor Paul’s argument is grounded in value judgments. The one part of his argument which might be a value judgment is setting an objective of social welfare maximisation, and you two seem to agree on that anyway. Paul is arguing from theory and evidence, as I think, are you. In principle you can test his claims. Therefore, not a value judgment. When you say you believe unemployment could cause a pareto-inferior equilibrium – I don’t think that’s a value judgment either. Its a guess that is testable, but you don’t have the evidence to hand. That doesn’t make it a value judgment. Its not as if the only option available is that you and Paul just agree to disagree. You’re not saying red is better than green. Even though it is 🙂
Fourth, I agree with your idea that it is dynamic view that matters. But there are at least two aspects of that idea you don’t explore. One is that while the present shock is temporary increases in government spending tend to be permanent. Even if government intervention is beneficial in the short term, a sustained increase in government’s share of expenditure of GDP will quickly overwhelm those benefits. Two, expansions in government expenditure arrive with lags. A recent stat doing the rounds is that less than (aprox) 10% of the $825 billion US package will be spent in 2009. The crisis may be over by the time the other 90% is used. And I’m sure you will agree there is no chance the US government will deploy those vast resources as productively as individuals spending their own money in an economy back on its feet.
Lots to weigh up.
Hi Matt,
“Matt I’m not persuaded by your position for a couple of reasons”
Very good – I appreciate critical analysis that is numbered 😉
“First, while the present crisis as a market failure is large I am not convinced it will be sustained in the same way that other alleged market failures (lemons, keyboards) can persist indefinitely, at least in theory. Doesn’t the “large, sustained” criterion you define here rule out a government response to the present crisis?”
This is an EXCELLENT question – and is something I should have already posted on but haven’t had the time. As I have said constantly, a permanent shock (such as permanently higher risk premiums) will lower potential output – and there is nothing the government can do to fix it. However, I had something else in mind when I mentioned a “sustained” externality.
Fundamentally, this can only be answered by steering to your second question
“Second, I’m not sure why you raise multiple equilibrium. Unless I’m mixing up the order of responses and replies, Paul has pointed out the evidence that economies seem to consistently converge on efficient solutions when multiple equilibria exist, and that it is hard to find actual examples of persistent market failures”
Multiple equilibria is the KEY to any belief in government action outside of a “temporary” shock. In the face of a temporary shock the government could “smooth” the economy in theory – but with the lags in fiscal policy this doesn’t seem like a good idea.
However, if the shock is sustained – but not necessarily permanent – interventionists have an argument.
In order to have a sustained shock that is not necessarily permanent, we need to have an economy that has multiple pareto ranked equilibrium.
Now, the idea of pareto ranked equilibrium is inherently difficult to test – as we cannot view where the economy converges given different initial conditions. However, we can observe the potential causes of multiple equilibrium.
I will talk more about this in the next comment …
Starting again – sorry had to get an ENORMOUS coffee …
One such example is an economy with increasing returns to scale in an economy. In the case of increasing returns to scale a short shock to general production can lead to a long-term decline in GDP and welfare in the economy – this is really a form of hysteresis.
Beyond this we have Keynes fundamental idea of “animal spirits”. If we are pushed sufficiently far out of equilibrium it is possible that a shift in expectations could lead us to converge to a completely different, pareto inferior, equilibrium.
Now, the concepts of increasing returns to scale, the persistence of “unemployment”, and the self-reinforcing nature of expectations are all testable. However, even if we tested and found them this does not give us incontrovertible evidence for multiple equilibrium 🙁
I would also add that there are two ways of viewing multiple pareto ranked equilibrium – at least in the way I see it. One set of people take it as an indication that the government can improve outcomes by redirecting resources (something the socialist calculation debate would have jumped on) – however, I don’t see government as knowledgeable or socially interested enough to push us to a better equilibrium in the case of full employment (at least given the current level of government spending and taxation).
Another view is that, in the face of a large shock the government can prevent the economy moving to a suboptimal solution. I think that as long as we see unemployment move sufficiently from its natural rate (the natural rate associated with what we believe is the pareto optimal equilibrium) then there is a case for government to do something – however, the something they do should be focused, rather than throwing money into the wind.
I will cover the final questions in another comment …
Hello again,
“Third, for what its worth, I’m quite sure neither yours nor Paul’s argument is grounded in value judgments”
Note that I am not using value judgment as a prerogative term – far from it, I respect it when people are transparent about their value judgments. However, in order to reach a conclusion about what is going on we have to make value judgments somewhere along the line. The evidence surrounding market failure, either for or against, is not water tight – and as a result we have to set our own weights on potential outcomes. These weights are the value judgments.
“Fourth, I agree with your idea that it is dynamic view that matters. But there are at least two aspects of that idea you don’t explore”
There are many dynamic elements I haven’t touched on 🙂
“One is that while the present shock is temporary increases in government spending tend to be permanent”
This is a relatively normative concern. Surely, if short-term stimulus was useful we could get the government to commit to keeping it short term in some way. Ultimately, if they don’t it is another case of government failure.
However, this is the sort of government failure that could be solve by economists actually being more active in the policy debate – if we inform the public and politicians maybe we could make a difference here …
“Two, expansions in government expenditure arrive with lags”
Definitely – in fact with infrastructure spending it is especially bad, as the projects aren’t on the go yet. That is why using government spending to smooth relatively short dips in economic activity, in the absense of hysterisis, doesn’t make much sense.
Fiscal tinkering is a long discredited way to run fiscal policy. If the government wants to do something it needs to ask:
1) What are the shocks to activity – are they structural?
2) Is activity likely to decline so sharply that we can get thrown into a suboptimal eqm?
In NZ I don’t believe we are facing such a case – 7% unemployment is not far enough away from neutral (value judgment 😛 ). In the US there is some heavy potential for them to get thrown into a suboptimal eqm – in which case there is a role for some government intervention.
However, they also have to keep in mind that potential output HAS FALLEN. When Krugman and co state that there will be a hughe “output gap” they are assuming that potential growth hasn’t been touched – a ridiculous assumption in the face of higher eqm interest rates.