I had the privilege of interviewing Peter Diamond, a professor at MIT and winner of the Nobel Memorial Prize in Economic Sciences in 2010. He is especially known for his analysis of US Social Security policy, serving as an advisor to the Advisory Council on Social Security in the late 1980s and 1990s. Nominated by Barack Obama in 2010 to serve on the Federal Reserve board of governors, he was blocked from the position by Senate Republicans.
Do you believe that market search costs are in structural decline as exponentially more potent technological advances, particularly in computer networking and data science, push traders closer and closer to perfect information?
No, I don’t. There’s no question that there are largely increased ways in which information is available. The critical question is how information feeds back into demand and so into pricing behaviour. I want to separate the consumer market from the labour market, which function differently – let’s start with the consumer market. Yes, some people can go online and use certain apps to find the best prices, but in the majority of cases that doesn’t happen. The people who have studied online pricing find the presence of the effects of limited knowledge of where to shop showing up in pricing distributions online. I don’t think that is going to change, although it may change in magnitude, but I doubt the extent to which mark-ups that are present and the note that distribution of mark-ups in equilibrium are changing in part because other things change as well, including if we focus on the US, the extent of concentration in the consumer goods market and obviously a lot of other markets as well. So I don’t think computers and the internet are changing the structure of retail pricing equilibrium.
For the labour market it’s quite a different story because the kind of modelling that was done when the labour market began was focused on the difficulty of locating people to consider hiring. The focus was then very much on small firms and not the large ones which have a standing pile of people who have applied and are available if the firms want them. And it also didn’t consider the kinds of coordinated markets like the one we’re going through right now, where there’s a certain extent of coordination in what’s going on, but very rarely fixed rules like those for example in the process of university admissions. We have different things in different markets, and if we go back to markets with small-firms, this looks a lot like it used to before – I live in a suburb, and when I walk downtown, and if things are like how they are now, there are stores with signs in their windows saying they’re hiring, with some information about the job. These are usually lower level jobs, but not completely. And secondly, by focusing on that aspect of the working of the labour market, what was left out is the process of evaluating candidates.
In the typical dynamic model, you wait to find a candidate, and then you instantaneously work out whether it’s a match or not, and what the wage is. And that leaves out the fact that particularly in a place where there are multiple applicants, there will be an evaluation process which will take significant time – it may take even more time than waiting for the right candidate. That happens too (more so in the US than the UK), for example at times of high unemployment, when there is high labour mobility, most of it from areas with low amounts of jobs to areas with high unemployment. This is inherently a time-consuming process. But for jobs which are posted online, where resumes and application information are posted online, where the firm picks a few applicants to follow up with further interviews, spread out over time, and have been left out in earlier modelling- models are generally built to try to understand one aspect of a function, but they only provide one part of the story. What is necessary to understand the entire process is to put together different components from specialised functions. A quote from Marshall: ‘The more you dig into a model, the more you’re learning, but the less it reflects reality.’ It’s a complex process, and some pieces are dramatically changed by the Internet, but many are not. The hiring process is still spread out in real time, which is the critical difference between an Arrow-Debreu world to the world we have today.
In 2012 you said: ‘This is no time for austerity – the federal debt level in the United States is still eminently manageable.’ Federal debt is set to reach $29 trillion in the near future. What should be done, if anything?
I think there are two things to think about. Obviously the US economy right now is reasonably strong, so there is room to do things with an eye further down the road rather than with a focus on the short term that you get with a depression. When you think about debt one of the concerns is that the debt market would start having fears about repayment ability or possible oncoming inflation. What’s important is whether a country’s public and private debt is denominated in its own currency or the currency of another country. US federal debt is obviously in US dollars, and the ability of the country to repay this debt, given that you don’t get the equivalent of a bank run, is very sound. So in my mind the first thought is that we don’t have an immediate crisis, but there is indeed a genuine issue, what is our policy on spending and taxes, and how are we going to organise those going forward to get the debt to an appropriate level given intergenerational and intertemporal issues?
My feeling is that the US since 1980 has spent far too little on investment in infrastructure, education, R&D – which we need more of and given how low interest rates are, borrowing for investment is a very sensible decision. On the other hand, my view is that what’s particularly important is the raising of a lot more tax revenue, which not only affects the debt level but also the politics of spending policies. If the need is for more investment, greater tax revenue will encourage the political process to make these kinds of decisions as well as helping with the intertemporal dimension. While I have no concern about short-term difficulties with paying the debt back, I have a lasting concern that the debt is negatively affecting the long-term economics and the short-term politics. More taxes would help in both cases.
For the United States specifically, given phenomena such as rising inequality and increased automation, is it time to begin severing the link between productive employment and the ability to afford necessary goods, perhaps through a Universal Basic Income?
Well, compared to most of the advanced European countries the US has a very weak safety net, so even without the issues of automation and the different functioning of various parts of the labour market, the US needs a much stronger safety net and a central issue is the extent to which the safety net should be tested and focused on the people who are relatively needy, or should be focused on universals, with an example being Social Security, which goes to everyone over 65 who has done enough work to qualify. In contrast, programs such as Medicaid and the SSI program for the poor and elderly are nowhere near as generous. Wilbur Cohen said that ‘a program for poor people is a poor program.’ So we’re discussing political economy as opposed to underlying normative economics. To my mind we need to do more with universal programs, and do more to support children, as opposed to focusing so narrowly on the children in the poorest families, and that involves direct cash transfers as well as better education and healthcare.
You have personally faced partisan, ideological barriers to implementing public policy, such as when your nomination to the Fed Board was blocked in 2010. Given the shifting political climate in many Western countries, do you think the prospects for well-informed, non-partisan economic policy is improving or worsening, and why?
That’s an easy question. I don’t think anyone would consider that it is improving. The concern here is the underlying political process which includes a big dose of the behaviour of voters as well as the behaviour of politicians and whatever rules get changed at the governmental level. That is showing, and I’ll focus on the US here, increasing difficulty in doing things that are bipartisan. There are many cases where there is enough public concern for a policy which fits within the stances of both parties, but parties are still more lockstep in the kinds of policies they are approving, much more so on the Republican side in my opinion, although this is happening on both sides. A huge change happened when the South went from reliably Democrat to reliably Republican, changing the degree of diversity in both parties.
So, the political process is a major concern and people are musing on things that can be done. Obviously change of personnel could have an impact, and a lot of thought is being given to changing rules at the local level, such as for example the possibility of ranked voting, which could have a significant impact on the political process. Experiments are being carried out at the moment on how voting takes place in primaries, and who can vote in those. The concern is very large, and it’s made even larger by the fact that we have some huge problems getting inadequate attention, starting of course with climate change, but also dealing with the various effects of computers and automation. All this calls for better quality politics: the issue is getting there.
How do you think these flaws manifest at the electoral level?
The American public and the American voters have a very negative attitude toward taxes. They don’t make a connection between taxes and gains from government spending, which is the case in for example some of the Scandinavian countries. Even tax increases that have sure-fire benefits for the public are viewed with great reluctance. For example, in Massachusetts the legislature discussed a small increase in the gasoline tax in order to primarily improve the road infrastructure, with serious road congestion problems and inadequate spending on public transport. It was voted down by the public.
The heart of this issue is the attitude of the public on the connection between the taxes they themselves pay and the benefits they themselves receive. There’s a quote from Earl Warren, former Chief Justice: ‘a program that benefits me is social progress, a program that benefits other people is socialism.’ Part of the problem is that the failure to make these connections runs through the political chain. Just because a tax is labelled as aid for the people who are paying it, it doesn’t mean that it stays this way. It can displace other spending that would have happened from other resources anyway, and it can spill over to be spent on other programs. What we need more of is taxes designed for a pre-designated purpose, with one example being Social Security, but others are quite rare. To my mind it is a very tough issue to get people aware of that link and get them to understand the benefits of higher taxes.