models

In 1991, the Soviet Union disintegrated. Once and for all, that demonstrated that capitalism is better than a “command economy,” at least in two aspects that mostly matter – productivity (effective use of resources) and ability to satisfy the needs and wants of consumers. Well, that conclusion was evident much earlier. Not many 20th century systems could be worse than the Soviet economy. (I know it first-hand. Prior to becoming a scientist, I worked for 10 years in the Soviet industry.) But does such a conclusion let one to rest one on his laurels? To maintain that you are the best, because you are better than Soviets? That was common consent of the “mainstream” American economists. As it turned out, a little premature. Superficial and short-witted. Capitalism is not a monolith. William Baumol and his co-authors in their 2007 book Good Capitalism, Bad Capitalism, and the Economics of Growth and Prosperity , define four types of capitalism: state-guided, oligarchic, big-firm, and entrepreneurial. The authors argue that the most productive is a mix of the last two types, that is, the combination characteristic for the American economy. That judgment may generally be true, but it seems far from universal. In today’s mixed economy, with dynamic relations between state and private sector, separating lines between them are often eroded. We see superb results of some state-guided industries and enterprises, too. South Korea has risen from nothing to a formidable economy because, in the 1960s and 1970s, it has undertaken exactly the same type of central-modeling-guided development (performed by the state together with a brilliant UN team of mostly American economists) that I recommend on this blog. Besides, productivity is not everything. One can pursue a wrong goal very productively. What should be the goal of that creative productive activity? Adam Smith viewed his individualistic approach just as a means, the end being the benefit of society. Like every living organism, a society has a goal – long-term sustainable survival in an acceptable state. From the societal point of view, all four defined above types of capitalism suffer from one common crucial drawback. If you don’t poke the nose of a capitalist enterprise into a negative “externality” (that is, an anti-societal side effect of its activities) and firmly hold it there, it will deny both the very existence and importance of that externality. Any enterprise wishes to get for free as much of limited societal resources as it can. (Here goes both the “invisible hand” of the market and the “visible hand” of the state.) The current state of capitalism is far from the best for society. Even in equilibrium, if the market prices do not reflect externalities and long-term goals of society, it does not provide a social optimum. “Optimum” could be declared as a rough approximation to reality two hundred years ago, when natural externalities were small in scale, gold standard prevented serious man-made externalities, and Great Britain and Europe were able to ship their unemployed to America and Australia. In this century, both Adam Smith and Milton Friedman are dead, wrong. Now we have chemical industry, zillion tons of carbon and sulfur emissions, and trillion-dollar deficits. To maintain the current-price market superiority under new conditions is a cruel hoax. An economic felony, shown as such in recent years. * * * Were we able to find the way to keep capitalists deeply aware of externalities, and block them from doing any societal harm, we could build a better capitalism – and perhaps prevent, mitigate, or postpone the catastrophes threatening us in this century. How then to do that blocking? Of course, there is criminalization of obviously harmful activities. There are also two other obvious routes: either to forbid or limit the volume of the harmful activities (regulation), or to change prices in such a way that the activity becomes disadvantageous and unprofitable. All three can also be combined. And how to find the proper levels of activity limits and price changes? Again, there are different ways to do that. The simplest way is to do what we are doing now — simply calculate some plausible numbers for quotas or taxes that would presumably counteract the first-order impacts of harmful activities. To influence natural externalities, like the effect of carbon emissions on the climate, we design “carbon tax” and climate bills. But externalities can also be man-made, such as trade deficits. “Compensated Free Trade” (CFT), described in my July 16 blog, is an attempt to influence a gross man-made externality, the trade deficit of the USA. For the latest 12 months (till April), this trade deficit equals $550 billion, or around 4 percent of GNP; that level is abnormally low because of the recession, when Americans have largely stopped “buying things they do not need with money they do not have to impress people they do not like.” At better times, the deficit did exceed $800 billion. But not to worry, there already was a deficit surge in recent months. Let me return to the crash of the Soviet Union. It turned out that it was triggered by exactly that deficit externality. We have to be eternally grateful to Ronald Reagan. He not only involved the Soviets in an arms race — he also persuaded Saudi Arabia to lower the oil price to 10 dollars per barrel. (Saudis did not lose: as a by-product, they destroyed the budding American artificial fuel industry. We need a similar Saudi action to deal with Iran, and I expect it will come soon. But beware sheikhs bearing gifts; let us not repeat past mistakes. Let us determine if we need such a clean-fuel industry and, if it is necessary, build it with government subsidies.) Before that, the foreign trade inflows and outflows of the USSR were approximately equal. The drop of oil price halved the inflows, so that the country started to borrow. In 2006, Yegor Gaydar, the deputy prime minister of Russia in the beginning of the 1990s, published a book Death of an Empire ; at the time, it was making tsunami waves in Moscow. He said that the Soviet Union disintegrated primarily because it had to borrow from its enemies. Caveat America! That was the second inference that could be made from the USSR crash. It was less evident but certainly much more valid than “our system is the best possible.” Also, it was more important. This conclusion was, however, very inconvenient for economists and politicians, therefore they didn’t do it. * * * I will not repeat here the arguments of my July 16 blog about the need for CFT. But in many decisions, in addition to the question “Why?” a second question arises: “Why now?” Because it is better to start from a low total USA deficit limit, which is possible only during a recession, thus discouraging hopes of the rest of the world for dipping again, after a recovery, in the fat wallets of those stupid Americans. Because we are starting a huge new stimulus, and, like the old one, it will end to a substantial degree in the deep offshore pockets (true, in deep Wall Street pockets, too). Ever tried to take out water out of a boat by a sieve? Because the total unemployment is 29.7 million Americans, or 18.5 percent of total workers ( Leo Hindery ). By another count, one out of four Americans is unemployed or underemployed ( Robert Reich ). We have to stop destroying our middle class. Because China, India, and Brazil will not enact any significant environmentalist measures (in spite of China now consuming more energy than the USA), and that will further increase their cost advantage over American enterprises. Because “Chimerica” is dead. What we see now are its death convulsions. American people say so. Congress feels it; the White House feels it. CFT is just a means to make the departure less painful — more gradual and comfortable for both parties. Because it is high time to restore the industrial might of this country, to return home our factories, and to make them competitive again, in spite of all unholy combinations of dirty tricks used by some of our trading partners. Because it is the very height of stupidity to behave like a sheep in a forest full of mercantilist wolves. (Even Paul Samuelson said so decades ago; Paul Krugman and, after him, Larry Summers just discovered that revelation and repeated it recently in their media articles. Sure, these poor babies did not know it before.) The sooner we stop coddling — just for sake of political correctness! — the unbalanced and dysfunctional global economic order, the better will be our chances of surviving the crisis. There are many more reasons, but I think that these seven are more than sufficient. * * * I think that everybody in their right mind understands that a global order based on persistent huge deficits of the USA is unsustainable. (Even if he does say the opposite. Tongue is given to man to conceal his thoughts. When one eminent diplomat died, another diplomat asked, “I wonder what he means by that?”) What we need is the fifth type of capitalism – a “balanced capitalism.” (Other possible names are “stabilized capitalism” and “stabilizing capitalism.” Please help to select the best term.) The simplest way to move toward it is as described above – by implementing measures that prevent or mitigate the first-order effects of negative externalities. But that is insufficient. As a rule, those measures underestimate the potential impact and the overall level of these externalities, when they are aggregated over the country or globally. (We need to “internalize” all countrywide externalities, natural and man-made, both short-term and long-term, as well as the global externalities that are related to the protection of global environment.) Besides, for the sake of generality, I will define as externality of social or economic decision not only its effect (“spillover”) on any party directly involved in that decision, as it is done usually. I will also include in this category any unforeseen consequence of that decision, even if it doesn’t impact a directly involved party. With sharply rising uncertainty, reaching now a radical, “uninsurable” level even for short-term, such “unforeseen” externalities become critically important. Moreover, it turns out that not one but rather two “invisible hands” are needed: one for the current production and one for developing new production capacity to satisfy the future demand. An enterprise demand includes covering the needs of the adjacent sectors of industry, so a producer has to have data about the future of not only his industry, but also all related industry sectors. (Since the market knows nothing about the future, the actual results are bad. This might have been acceptable in earlier, more tolerable times, but not in time of sharply reduced domestic investment and economic crises, when “animal spirits” are close to a zero level.) We have to go beyond the first-order effects, too. All economy sectors are ultimately interconnected, and we actually need the so-called “input-output analysis” table, describing the whole economy in terms of coefficients of intersector relations, possibly with some geographical aspects included. The past values of these coefficients are available at many universities and government organizations. But we need their future values, the values after the future innovations, which may or may not occur. We also need to take into consideration the long-term survival goal of the society, represented by its approximate proxies. As such proxies, short-term objectives and constraints include: sufficient growth, low unemployment, stable prices, sufficient satisfaction of needs and wants of population, and a healthy balance of payments. Main long-term ones are: preservation of the industrial base, preservation of the middle class, and attainment of social and geopolitical goals of the country. Of course, all these data would be very rough approximations: we know almost nothing about the future. At best, we can get no more that “guesstimates.” We can make some plausible predictions about the technological trends, but almost nothing is known about connections of the above proxies to the ultimate goal of long-term survival of society. To get all these data would take years of superhuman and costly efforts of a substantial group of scientists. But suppose that we obtained the data. What do we do with it? What results could justify such a Herculean endeavor? A multi-scenario optimization model under radical uncertainty naturally suggests itself. I will not talk in this blog about such a model and methods of its solution. (My system of Risk-Constrained Optimizationâ, or RCO, specially designed for solving such models, will be outlined in the next blog.) Here I will say only that if we could formulate, fill with data, and solve such a model, we could obtain from it the so-called “dual prices,” which would replace the today’s deceptively good prices. The new prices will “nudge” the decisions in desirable direction. If we want to preserve cherry orchards, we should include that request in the model, and it will tell us the required price of cherries. Then we can decide – is our wish affordable, should we abandon it or subsidize it via non-market channels? A capitalism using such prices (mind you – not commands, just different prices) would probably be as close to societal perfection as humanly possible, while not losing any of its positive traits. Moreover, it will be more productive, too, because the information about the future, to be generated by the model, would help private enterprises. Of course, this information will be very far from perfect, but it still should be much better than what they have now. Most important, it will contain data about potential scenarios, their risks and opportunities, and the technologies and strategies best suited to each scenario. There undoubtedly exists a pressing need for society-wide modeling. It can be done, too – the experience of South Korea, cited in the beginning of this blog, as well as my own experience (both are outlined in the next section) confirm that. Of course, the economy of the USA is not the economy of South Korea or the USSR, and planning under radical uncertainty is in some respects a couple of orders of magnitude more difficult than deterministic planning (it reduces though the requirements to quality of data), but the scientific and computational resources of this country are incomparably better, too. The work can be started on small scale, with a very aggregated model, and gradually expanded and detailed. Although the collection of sensitive technological data can probably be trusted only to a government organization, a university or other scientific organization can initiate preliminary methodological work. In time, that will bring substantial benefits to the entity, too. Tennis, anyone? * * * This section is intended primarily for economists. Of course, it contains some interesting non-technical information, too, and non-economists are welcome to browse through it. I’d be happy to provide additional explanations. In South Korea, great attention was given to the technological changes due to new investment. Industry teams of specialists were formed to estimate the changes to be expected from new capacity coming on line, for each year of the planning period. In that way, dynamic (year-by-year) input-output coefficients were generated for all sectors of the economy. In addition, overall constraints were imposed on balance-of-payment deficit, unemployment, minimal school attendance, and several society-wide externalities. South Korea didn’t even have computers yet. The computational work of the input-output analysis (inversion of the matrix by an inferior pivot method, necessary because of that absence) was performed by thousands of people, for months using abacuses or turning the handles of primitive mechanical calculators. Amazingly, these efforts were so well coordinated and executed that they provided highly precise results. Interestingly, post-mortem calculations performed after the planning period had shown that the model forecasts would have come closer to actual outcomes, if static (one-period), rather than dynamic model were used. (See below about the use of a static version in my 1965 model.) The team members received orders of honor, specially established by the Korean government to celebrate exceeding the targets of the Second Five-Year Plan developed by the state on the UN team recommendations. Irma Adelman, a prominent American economist, one of the leaders of the team, proudly told me the details of that project. It was an undeservedly forgotten tour de force, perhaps one of the highest practical achievements of the “dismal science.” In my opinion, if anything in economics ever deserved a Nobel, that was it. Not Utopian concoctions with mathematical fig leaves. As to my own experience, it was as follows. Today, in 2010, I am terribly worried about the current market prices that can easily lead to various global and American catastrophes. In 1964, I was similarly worried about prices of Soviet command economy. They were ridiculous: for obvious political reasons, bread (for instance) was made super-cheap, and all trains from Moscow were full of peasants, each of them bringing home dozens of bread loaves bought in the city, to feed to their cattle and pigs. I had just been appointed to head the laboratory of long-range planning of sectors of industry at CEMI, a Moscow think tank (see more details in my bio). But I could not in good conscience optimize anything under such made-up prices. Therefore I had to go out of my sector-of-industry box and to do something about the whole economy. Very reluctantly, because society-wide modeling is incomparably more complex and difficult than modeling an industry sector. So in 1964-1965, I developed a model for long-range planning of the Soviet economy by industries and regions. (Thankfully, I was guided mostly not by economic “science” but by common sense.) By the way, I also (similar to the South Korean team) had imposed in that model the constraints on the balance-of-payments deficit. Saves from a crisis, you know. Pity it is not done in this country now. (My 1965 and 1967 articles about that model, in Russian and English, respectively, are available at my website .) For several years, the model was a banner project of CEMI. By a government decree, 400 planning and research organizations provided the information for the model. These organizations forecasted technological change, defined trends in consumption and interregional migration, quantified externalities, and so on. Two CEMI laboratories with staff of about 70 people were set up to process that information. I had developed not only the model, but also the algorithm for solving the enormous non-linear programming problem arising from that model. Several techniques were combined to simplify the original model, reduce its dimensions, decompose it into a number of small subproblems, and replace some parts of non-linear programming by matrix inversion methods. One of those techniques was transformation of a dynamic year-by-year model for a ten-year planning period into a static model. As found in South Korea, that might in some cases even improve the results. My laboratory implemented those model and algorithm. The original problem had millions constraints and scores of millions of variables. After all transformations, it was successfully solved by 1972 on computers able to handle much simpler linear programming models with only up to 400 constraints. As far as I know, the results were not implemented. By that time, I had already applied for emigration and was “persona non grata.” * * * The last week exchange of comments to my July 16 blog “Compensated Free Trade” brought up an important point that might be of interest to the present readers. One commentator has suggested that the system of balancing foreign trade, proposed in 2003 by Warren Buffett, is better than CFT. In that system each American exporter receives certificates in the amount equal to the value of his exports. He can auction those certificates to the would-be exporters to the USA, who thus acquire rights of exporting to America of goods that would bring them the same amount of money. Of course, the WB system would very soon achieve a trade balance. But, because the WB system does not set deficit limits for individual countries, as CFT does, is has one vital flaw. A trading partner that has substantial dollar reserves can overbid his competitors and buy all existing certificates. He can use those of them that he needs for his own exports, obtaining (in addition to his presumable cost advantage) another advantage, and becoming as close to an export monopolist as he wishes. (In that process, he can easily destroy his competitors.) He can then sell the rest of certificates to other exporters at somewhat reduced prices, possibly using the certificates as tools of political or economic blackmail. But, even without knowing about that flaw (which I discovered later), Paul Samuelson wrote to me in his letter of October 14, 2004: “I think that Buffett goal would have strong consequences, probably more bad than good.” After a paragraph describing his simulation of the WB system in his models, he continues: “By 2020, the post-Buffet U.S. would have fallen behind the EU and the Pacific Rim. If you disagree, you may well turn out to be right. In economics, 2 + 2 = 4 arguments can rarely settle practical problems. Good luck, Paul Samuelson” Understandably, I was very happy that Samuelson did not find any flaws in my CFT proposals; that he seemingly preferred CFT to the WB system; and that (as told by his assistant) he kept my letter and my article on his desk for unusually long time. Copyright © 2010