Eddie asked two interesting questions in yesterday's post which time pressure enabled me to conveniently duck, but which I can't really avoid forever. The first is much easier:
When you talk about "labour market reform has a similar look to it..... This writing down takes two forms. First lower salary expectations, and secondly lower compensation when jobs are 'bought out'." Again, I share these sentiments. But what do you mean by bought out? I guess labour market reforms essentially translate into releasing more labour into the market …
The essential point here is to do with corporate 'restructuring', 'downsizing' and 'outsourcing'. Now the labour reforms here in Europe are often related to reducing the cost to the company off taking workers of the books. That in turn implies that the workers receives less financial compensation when they go. Now for many years there has been an implicit contract between company and worker that the worker remained with the company, and gave their accumulated experience at what may not have been market price in return for job security and anticipated compensation when the time for 'early retirement' came. Now it is this experience that, as a result of technological change and the changing international structure of production, no longer has the value it once had. So it is logical that the 'buyout' payment should be less. What I am saying is that the market mechanism had given this value yesterday to the worker's service and experience, and today, for the above mentioned reasons, the market puts another value. This is tantamount to devaluation, devaluation of the worth of the job buyout. This depreciation is every bit as important for the economic structure as is asset price deflation, but receives much less attention. This same worker is now to be hit twice, since later the 'pensions' reforms will reduce the other part of their implicit savings: this is a double whammy which is every bit as important in its way as the money lost in the Argentine bank accounts. The market value of your life savings just changed. Possibly this value 'redistribution' might not be so problematic from a technical economic point of view if it were simply a transfer of values downwards to younger generations. But that is the whole point, the younger generations are much smaller. So consumption takes a hit, and private saving has to go up. Go up that is until it can hold no more: in Japan private saving is now declining steadily. in other words, there may well be a point where the process turns back in on itself again, an inflection point.
Now for the second, and trickier, argument
“the idea of 'potential output' may be useful as a rule of thumb guide, but it becomes deeply flawed - for the above mentioned creative destruction reasons - when you want a richer more dynamically oriented analysis: the type of analysis for which neo-classical economics is relatively poorly equipped”
Look, I would like to make a couple of preliminary points here. I have a tendency to think out loud, and to do it in ways which is sometimes annoying. What is he talking about? I think blogging is about sharing even half-thoughts with others and then in the ensuing debate sorting out the details on the fly. I'm sure this way of thinking and doing will irritate some. What it means is that sometimes I say things which to me seem obvious, but which are much more difficult to justify. Sometimes these things are provocative, sometimes they are wrong. I hope I will have the courage to admit to the 'bummers'. What we are into here, I think, is the Beta 1.0 version release of ideas. This is a bit different to the way things have been, but I think it is the way we should go. I think the Linus Thorvald project management model has much more general applicability than is normally appreciated. We are all each other's eyes, ears, and neuronal bridges. It is risky, but it may be justified since we are facing so much that is new, and if we are to generate appropriate responses we need to think quickly. Brad wrote an early internet paper called Tools For Thought. I think it is one of his most important papers. What we have here is a communications-based conceptual revolution. Like all the rest of them (speech, writing, printing) the internet is producing news ways of thinking and new ways of thinking about thinking. Nothing less.
The latest example of what I am trying to get away from came in the mail this morning. A greek anthropologist, who I had written to about putting one of her paper's - on Graeco-Bulgarian Pomacs - on my website (which I think was offering a favour) replied saying she was awaiting reference revisions prior to publication, so could I please wait for the final version. While I respect this colleague, and appreciate her point of view, I think it is something we should be trying to get away from. What are we afraid of, or embarrased about here. As Maynard noted in an earlier post, the Emperor himself has no clothes and doesn't give a damn! Why the hell should we ordinary mortals be worried. Publish early and publish often!!
Now, this preamble was not a way of avoiding a difficult question, just a methodological footnote. My essential argument is that sometimes we have a bit too much Keynes, and not enough Schumpeter. The two are perfectly compatible, and a good example of this can be found here . This book is one everyone who wants to understand economics should read and thoroughly assimilate. Now one of Christopher Dow's main points of departure for looking at the limitations of neo-classical orthodoxy is what he calls the competitive equilibrium path. An associated concept, and one from which the output gap idea seems to be derived, is the potential output path. The point about both of these terms is that they are abstractions. Other abstractions are potentail output and full employment. We blind ourselves when we think of these as physical attributes, so many people working or so many factories producing so many products. At what level of labour market participation is the economy running at 'full employment'? Well to answer that you need to resort to the potential output path. And potential output is expressed in prices (whether nominal or real) and not in widgets. And to discuss both of the above meaningfully you need to resort to expectations. And from here on things get more complicated.
Now this path is in fact a problematic concept. One of the reasons it is problematic is that any dynamic economy is in a constant process of re-evaluation of its component a parts. This is what I am trying to argue about human resource accumulation above. Your experinece has one value now, and another later, depending. The same is also true of embodied capital, which can be written off this way now, and that way later, depending. And all this problematic writing down hasn't come to an end with the internet boom. We are still in a time of fundamental, destabalising, point-of-view-changing tecnological movement - in fact it looks like we always will be. Now to calculate a time path for potential output you need a time path for embodied capital and a time path for human capital. Well this, in my book, is where the curve (and my brain) starts bending. Since all the values are related to other values in a seemingly endless regression, my feeling is it is impossible in principle to construct such time paths. This is the point about strategic and fundamental uncertainty and the lack of visibility. We need to model better technological changes, and better understand ideas like TFP. This is work in progress. There are no results or definitive answers, only questions.
Meantime, I have a confession. I also use the output gap argument. It is one of those convenient fictions (like Okun's law) that enables us to talk about things and explain them, in just the same way that ther was a time when I told my children that you could not divide anything by zero. So long as we aware of the limitations, where is the harm? Over a single time period the rule gives an accurate enough result. If in this time period potential output grows more rapidly than GDP there will be price deflation.