WEBVTT NOTE Milton Friedman versus the Rothbardians: An Empirical Test 1 00:00:00.000 --> 00:00:08.600 Basically, in the last 50 years since the publication of Rothbard's formulation of a business cycle theory and his analysis of the Great Depression, 2 00:00:08.600 --> 00:00:22.100 there's been a really amounting recognition and respect from mainstream economists who are increasingly challenging this sort of orthodox Friedman-Schwartz explanation of the Great Depression. 3 00:00:22.100 --> 00:00:27.100 We see this in a couple of papers, for example, in the Journal of Political Economy, 4 00:00:27.100 --> 00:00:34.100 Ohanian and Cole challenged the idea that the New Deal had anything to do with getting us out of the Depression. 5 00:00:34.100 --> 00:00:39.100 And then later on, in the Journal of Economic Theory, another very high-level journal, 6 00:00:39.100 --> 00:00:45.100 Ohanian alone wrote an article called, Who or What Caused the Great Depression? 7 00:00:45.100 --> 00:00:50.100 And he cited Rothbard, he took the Rothbardian view that, in fact, 8 00:00:50.100 --> 00:00:53.100 It was rigidity in labor markets that caused it. 9 00:00:53.100 --> 00:00:57.100 He explicitly rejected the Friedman-Schwartz view that it was a monetary failure. 10 00:00:57.100 --> 00:01:04.100 He said it was a failure of labor markets, not a failure of the Fed in this particular case. 11 00:01:04.100 --> 00:01:11.100 And then, of course, we know that many financial economists, practitioners, commentators 12 00:01:11.100 --> 00:01:18.540 have also been using the Rothbardian formulation of Austrian Business 13 00:01:18.540 --> 00:01:24.100 Cycle Theory going forward in the last decade or more to explain events as they 14 00:01:24.100 --> 00:01:28.100 unfolded leading up to the bursting of the housing bubble and then the 15 00:01:28.100 --> 00:01:34.860 financial meltdown. So in the first part of my paper I go through a number 16 00:01:34.860 --> 00:01:39.540 of different people I've used Rothbard. In fact I'm Brendan Brown who's here with 17 00:01:39.540 --> 00:01:44.540 with us also use Rothbard's America's Great Depression 18 00:01:45.460 --> 00:01:50.460 in some of his analysis of the Great Depression 19 00:01:51.220 --> 00:01:56.220 in his book, The Global Curse of the Federal Reserve. 20 00:01:56.660 --> 00:02:00.860 So I found the paradox and that was that 21 00:02:00.860 --> 00:02:03.200 some contemporary Austrian economists 22 00:02:03.200 --> 00:02:04.980 have recently criticized Rothbard 23 00:02:04.980 --> 00:02:08.460 as a greatly inferior monetary economist to Friedman. 24 00:02:08.460 --> 00:02:38.460 Let's take one example, George Selgin, and I'm not completely being fair to George here, but he's never fair to me, so, because it was on a blog, he's a free banker who has strong Austrian inclinations, he harshly compared Rothbard to Friedman in an online article blog in 2011, and he said as a monetary economist Rothbard was mediocre to bad, his version of the Austrian cycle theory was naive, containing about a dozen auxiliary 25 00:02:38.460 --> 00:02:41.460 all of which are patently false. 26 00:02:41.460 --> 00:02:46.460 Milton will always seem to me the bigger man as well as the better monetary economist. 27 00:02:46.460 --> 00:02:50.460 So I didn't want to really refute that. I wanted to have some fun with it. 28 00:02:50.460 --> 00:02:54.460 And refuted in sort of a different way. 29 00:02:54.460 --> 00:02:58.460 And it struck me that you could really test Selgen's assertion. 30 00:02:58.460 --> 00:03:02.460 Now, the test I proposed to conduct, or that I did conduct, 31 00:03:02.460 --> 00:03:06.460 does not really involve specific quantitative predictions 32 00:03:06.460 --> 00:03:14.460 Predictions 33 00:03:36.460 --> 00:03:42.020 People who pretty closely follow him, including myself, Mark Thornton, and there are others, 34 00:03:42.020 --> 00:03:48.060 made predictions about what the pattern of the unfolding of these events would look like. 35 00:03:48.060 --> 00:03:53.660 And that's what I did then was to talk about pattern prediction. 36 00:03:53.660 --> 00:03:55.260 So let me just give you a little bit about pattern prediction. 37 00:03:55.260 --> 00:04:01.900 Now this concept was made famous by Friedrich Hayek, okay. 38 00:04:01.900 --> 00:04:08.500 So I went back and I began to look and it turned out that it was actually first articulated 39 00:04:08.500 --> 00:04:14.780 about 30 years before Hayek wrote by Lionel Robbins in his brilliant essay on the nature 40 00:04:14.780 --> 00:04:18.060 and significance of economic science. 41 00:04:18.060 --> 00:04:21.860 He really first formulated the concept and he actually used the terminology of pattern 42 00:04:21.860 --> 00:04:25.300 prediction. 43 00:04:25.300 --> 00:04:28.700 Let me just give you a few quotes from Robbins just to give you a flavor of what pattern predictions 44 00:04:28.700 --> 00:04:35.220 If the given situation conforms to a certain pattern, certain other features must also 45 00:04:35.220 --> 00:04:40.820 be present, for their presence must be deducible from the pattern originally postulated. 46 00:04:40.820 --> 00:04:46.940 The analytical method is simply a way of discovering the necessary consequences of complex collocations 47 00:04:46.940 --> 00:04:54.340 of facts, consequences whose counterpart in reality is not so immediately discernible 48 00:04:54.340 --> 00:04:56.900 as a counterpart of the original postulates. 49 00:04:56.980 --> 00:05:01.180 Granted the correspondence of its original assumptions and the facts, 50 00:05:01.260 --> 00:05:04.020 its conclusions are inevitable and inescapable. 51 00:05:04.100 --> 00:05:07.780 It is this inevitability of economic analysis 52 00:05:07.860 --> 00:05:11.980 that gives it its very considerable prognostic value. 53 00:05:12.060 --> 00:05:13.380 Those are his words. 54 00:05:13.460 --> 00:05:17.300 Robbins went on to point out that while economics cannot predict changes 55 00:05:17.380 --> 00:05:19.700 in valuations and technical facts, 56 00:05:19.780 --> 00:05:23.540 once the pattern of such data has been ascertained, 57 00:05:23.540 --> 00:05:36.540 It can draw inevitable conclusions as to their implications, and if the data remain unchanged, it can draw inevitable conclusions as to their implications. 58 00:05:36.540 --> 00:05:45.540 Now Robinson can see that when there's a very small change in the data, an exogenous change, a tariff on one single item in the economy, 59 00:05:45.540 --> 00:06:01.540 that can certainly not show up in the later data, that can certainly be hidden by other exogenous changes that may take place in the interim before the full consequences of the tariff have unfolded. 60 00:06:01.540 --> 00:06:09.540 And he said, for example, wages could have fallen, in which case the price of the article that had a tariff placed on it may not actually show up in the data. 61 00:06:09.540 --> 00:06:19.540 However, he went on to something very interesting. He says there are certain broad changes, usually involving many lines of expenditure or production at once. 62 00:06:19.540 --> 00:06:26.540 Where a knowledge of implications is a very firm basis for conjectures of strong probability. 63 00:06:26.540 --> 00:06:29.540 This is particularly the case in the sphere of monetary phenomena. 64 00:06:29.540 --> 00:06:34.540 According to Robbins, pattern predictions relating to the later phase of business cycle, 65 00:06:34.540 --> 00:06:40.480 cycle, based on data revealing that the first phase is already in progress, falls into the 66 00:06:40.480 --> 00:06:46.100 category of what he called conjectures of strong probability, explained Robbins. It 67 00:06:46.100 --> 00:06:51.340 becomes more and more clear, for purely analytical reasons, that once the signs of a major boom 68 00:06:51.340 --> 00:06:56.700 in trade have made their appearance, the coming of slump and depression is almost certain. 69 00:06:56.700 --> 00:07:00.600 So now he's saying that this is not a Ceteris Paribus prediction. I mean, this is a prediction 70 00:07:00.600 --> 00:07:06.600 in the data. Although when it will come and how long it will last are not matters which 71 00:07:06.600 --> 00:07:12.200 are predictable since they depend on human volitions appearing after the indications 72 00:07:12.200 --> 00:07:21.200 and questions have appeared. Mises sharpened Robin's formulation of the concept of pattern 73 00:07:21.200 --> 00:07:26.280 prediction. According to Mises, economics provides the indispensable means of ascertaining 74 00:07:26.280 --> 00:07:31.240 whether government interventions of various types have succeeded in achieving the goals 75 00:07:31.240 --> 00:07:32.740 they were aiming at. 76 00:07:32.740 --> 00:07:37.080 That is, whether the predictions of the supporters of the interventions have come to pass. 77 00:07:37.080 --> 00:07:42.120 For the aims of the interventionists, you know, it surprised me in thinking this through, 78 00:07:42.120 --> 00:07:47.560 are invariably articulated as a specific pattern of economic activity that will diverge from 79 00:07:47.560 --> 00:07:50.520 that which the unhampered market would bring about. 80 00:07:50.520 --> 00:07:54.600 All interventionist measures are therefore based on implicit pattern predictions. 81 00:07:54.600 --> 00:08:01.600 Now, Mises doesn't say that in those words, those are my words, but he has an insight into that. 82 00:08:01.600 --> 00:08:17.600 So, in tracing out the effects of any actual government intervention, therefore, the deductive economist is in effect enunciating a prediction of an alternative pattern of economic interactions that will inevitably emerge once the interventionist policy has been implemented. 83 00:08:17.600 --> 00:08:26.600 And Mises goes on to say that because of human valuations and volitions, these can only be qualitative, not quantitative. 84 00:08:26.600 --> 00:08:30.600 So he forcefully states the claim for the predictive value of economic science. 85 00:08:30.600 --> 00:08:37.600 Quote, economics can predict the effects to be expected from resorting to definite measures of economic policies. 86 00:08:37.600 --> 00:08:42.600 It can answer the question of whether a definite policy is able to attain the ends aimed at. 87 00:08:42.600 --> 00:08:46.600 And if the answer is in the negative, what its real effects will be. 88 00:08:46.600 --> 00:08:55.600 But of course, this prediction can only be qualitative, it cannot be quantitative, because there are no constant relations between the factors and the effects concerned. 89 00:08:55.600 --> 00:09:04.600 The practical value of economics is to be seen in this neatly circumscribed power of predicting the outcome of definite measures. 90 00:09:04.600 --> 00:09:12.600 Mises is saying, this is not a Cerebus Paribus prediction, this is a prediction that you will see configured in the data. 91 00:09:12.600 --> 00:09:22.760 In fact, Mises goes a little bit further than that and he says, I'm paraphrasing him, for 92 00:09:22.760 --> 00:09:27.880 Mises then, pattern predictions refer to qualitative configuration of objective economic data which 93 00:09:27.880 --> 00:09:32.520 could not be detected without prior economic analysis and indeed whose very conception 94 00:09:32.520 --> 00:09:34.000 is dependent on economic theory. 95 00:09:34.000 --> 00:09:35.800 Let me give you an example. 96 00:09:35.800 --> 00:09:40.860 An increase in the federal minimum wage to $9, which is now being bandied about per hour 97 00:09:40.860 --> 00:09:45.860 or from the current $7.25 per hour which we assume to be above the equilibrium wage will 98 00:09:45.860 --> 00:09:50.300 lead to an increased surplus of unskilled labor reflected in greater unemployment or 99 00:09:50.300 --> 00:09:56.380 decrease in its non-wage benefits or both. This pattern prediction is theory-dependent 100 00:09:56.380 --> 00:10:03.340 in the sense that the very phenomenon of an excess supply of labor specified by the theory 101 00:10:03.340 --> 00:10:07.700 could not be even apprehended without it, okay, is specified by the theory and could 102 00:10:07.700 --> 00:10:12.500 would not even be apprehended without it, let alone detected in the complex objective data. 103 00:10:12.500 --> 00:10:21.100 So the very prediction of the pattern, you can't even know what the pattern would be like without prior analysis. 104 00:10:21.100 --> 00:10:24.100 So the empirical stuff comes afterwards. 105 00:10:24.100 --> 00:10:27.700 Okay, let's get to Friedman. 106 00:10:27.700 --> 00:10:34.700 Friedman treated the relationship between money and prices as aggregative and mechanical. 107 00:10:34.700 --> 00:10:49.700 He focused narrowly on the effect that a change in total quantity of money would have on a unitary variable that he called the price level, which can be measured in a myriad of ways, which we won't get into. 108 00:10:49.700 --> 00:11:00.700 But this yields a definition of inflation that refers exclusively to the general movements of prices and completely ignores other more important effects of changes in the money supply. 109 00:11:00.700 --> 00:11:14.700 These effects include artificially depressed interest rates, a distorted structure of relative prices, the falsification of entrepreneurs, profit and wealth calculations, all of which lead to asset bubbles, malinvestments, overconsumption and financial crises. 110 00:11:14.700 --> 00:11:21.700 So because of Friedman's defective analytical economics, his pattern predictions are going to be incorrect. 111 00:11:21.700 --> 00:11:29.700 So what I did was go back to 2002 and found articles written by Milton Friedman and he was still very lucid at that time. 112 00:11:29.700 --> 00:11:38.940 at that time, in every year from 2002 to 2006, and even one that was published posthumously 113 00:11:38.940 --> 00:11:48.380 the day after he passed away. And he completely misses the whole pattern of the bubble and 114 00:11:48.380 --> 00:11:54.280 the housing bubble, the run-up in financial prices and so on. He thinks the economy is 115 00:11:54.280 --> 00:12:19.280 Now, in my paper, you know, I have, as I said, one or two articles in which he makes these claims for each year, but let me skip, in the interest of time, to a remarkable interview that he gave to Charlie Rose on public television in December of 2005, and let me quote from that. 116 00:12:19.280 --> 00:12:29.600 covered a broad range of topics including the performance of the Fed under Alan Greenspan. 117 00:12:29.600 --> 00:12:34.840 Friedman gave no indication even at this late date that he had a clue that monetary policy 118 00:12:34.840 --> 00:12:39.880 had stimulated a dangerous housing bubble or that a bubble even existed. 119 00:12:39.880 --> 00:12:45.400 By this time the money supply had expanded by two trillion dollars or about one billion 120 00:12:45.400 --> 00:13:09.400 Friedman effusively praised Alan Greenspan and gave his approval to the Fed's recent 121 00:13:09.400 --> 00:13:11.000 monetary policy. 122 00:13:11.000 --> 00:13:16.040 The interview is worth quoting at length because it demonstrates how very bad monetary theory, 123 00:13:16.040 --> 00:13:22.400 how bad analysis, gives rise to wildly inaccurate interpretations of the economy's performance. 124 00:13:22.400 --> 00:13:28.840 Not just interpretations, but pattern predictions. Note how Friedman is almost monomaniacal in 125 00:13:28.840 --> 00:13:33.840 his focus on price stability, in what I'm about to read, and how this drives his entire 126 00:13:33.840 --> 00:13:39.720 appraisal of the state of the U.S. economy. Okay, so now I'm going to quote Milton Friedman. 127 00:13:39.720 --> 00:13:45.720 The United States is at the peak of its performance in its history. There has never been a time 128 00:13:45.720 --> 00:13:51.200 in the United States when we have had the state of prosperity, its level and its spread 129 00:13:51.200 --> 00:13:55.680 that we have had in the last ten or fifteen years. There has never been a fifteen year 130 00:13:55.680 --> 00:14:01.560 period in which there has been so little fluctuation in prices and inflation. Inflation has stayed 131 00:14:01.560 --> 00:14:05.520 around two or three percent or less for the last fifteen years. I mean this is it, that's 132 00:14:05.520 --> 00:14:12.040 It's unprecedented. I certainly do give credit to Alan Greenspan for that. I think monetary 133 00:14:12.040 --> 00:14:17.960 policy is primarily responsible for it. Charlie Rose, you think that Alan Greenspan was the 134 00:14:17.960 --> 00:14:23.480 greatest Federal Reserve chairman ever? I don't know if that was the inflection but 135 00:14:23.480 --> 00:14:29.640 I like to make it. Milton Friedman, there has been no chairman since the founding of 136 00:14:29.640 --> 00:14:34.880 the Fed who has anything like as good an outcome. Because he took the containing of inflation 137 00:14:34.880 --> 00:14:38.760 as a Chief Task of the Fed. 138 00:14:38.760 --> 00:14:39.760 Let me put it this way. 139 00:14:39.760 --> 00:14:44.560 In the first 75 years of its existence, the Fed on the average was a major negative feature 140 00:14:44.560 --> 00:14:45.840 in the economy. 141 00:14:45.840 --> 00:14:49.440 We never would have had the Great Depression if there hadn't been a Fed. 142 00:14:49.440 --> 00:14:55.560 Since then, since 1982 or 1983, the Fed has been a beneficiary, and I mean to say benefactor 143 00:14:55.560 --> 00:14:58.520 there, for the economy. 144 00:14:58.520 --> 00:15:03.240 So that I had never, or I had very seldom, anything good to say about the Fed before 145 00:15:03.240 --> 00:15:04.520 the 1980s. 146 00:15:04.520 --> 00:15:08.840 Since Alan Greenspan took over, I have very little but good to say, and by the way in 147 00:15:08.840 --> 00:15:17.580 some of the earlier articles that I looked at between 2004 and 2002 and 2005, Friedman 148 00:15:17.580 --> 00:15:22.560 goes on, makes comments that we don't even need his quantity rule anymore. 149 00:15:22.560 --> 00:15:27.660 He actually abandons, he forsakes his quantity rule because the Fed has now under Alan Greenspan 150 00:15:27.660 --> 00:15:34.000 figured out the correct thermostat by which to adjust changes in velocity, to adjust the 151 00:15:34.000 --> 00:15:35.800 The Money Supply to Chains and Velocity. 152 00:15:37.680 --> 00:15:40.700 So let me continue with the interview. 153 00:15:46.640 --> 00:15:49.600 I think his successor is a very able man, 154 00:15:49.600 --> 00:15:53.620 and he, like Greenspan, takes keeping stable prices 155 00:15:53.620 --> 00:15:55.440 as the major function of the Fed. 156 00:15:55.440 --> 00:15:56.920 And I have a good deal of confidence 157 00:15:56.920 --> 00:15:59.640 that he will continue in Alan Greenspan's path. 158 00:15:59.640 --> 00:16:00.840 You can make a good argument 159 00:16:00.840 --> 00:16:03.640 that the Fed overdid easing money a little, 160 00:16:03.640 --> 00:16:10.200 and that must be a reference to when they push rates down to one or two percent. 161 00:16:10.200 --> 00:16:11.600 One percent. 162 00:16:11.600 --> 00:16:17.080 That they kept the fed funds rate at one or two percent a little too long 163 00:16:17.080 --> 00:16:20.480 and you know that it's a natural tendency that you overdo things. 164 00:16:20.480 --> 00:16:27.080 You almost never go right along on a stable path like this and he makes his hand go straight up. 165 00:16:27.080 --> 00:16:29.200 He says it waves a little bit. 166 00:16:29.200 --> 00:16:33.000 He says you go up, you go too far and it's a little hard to calibrate 167 00:16:33.000 --> 00:16:37.080 but that was Greenspan's genius. He could calibrate everything. He was willing to 168 00:16:37.080 --> 00:16:40.340 abandon the quantity rule for the Greenspan rule. 169 00:16:40.340 --> 00:16:49.200 Posthumously, the day after he died, November 17, 2006, an article that he had 170 00:16:49.200 --> 00:16:51.080 written came out in the Wall Street Journal, 171 00:16:51.080 --> 00:16:55.320 Why Money Matters. In this, Friedman compared three episodes of monetary 172 00:16:55.320 --> 00:17:02.340 policy which he called, using positivist jargon, a major natural experiment. And the 173 00:17:02.340 --> 00:17:07.820 The natural experiment was to compare the United States economy in the 1920s, the Japanese 174 00:17:07.820 --> 00:17:18.620 economy in the 1990s, and then the U.S. economy in the 1990s also, when we had the tech bubble. 175 00:17:18.620 --> 00:17:23.980 So what he says is, the conclusion that Friedman drew from this natural experiment was that 176 00:17:23.980 --> 00:17:30.260 the highly expansionary monetary policy that the Fed pursued after the 1990s boom caused 177 00:17:30.260 --> 00:17:44.260 allowed the U.S. recession of 2000-2001 to be very mild and allowed the U.S. economy to avoid a 1930s style Great Depression or a 1990s style Japanese Great Recession. 178 00:17:44.260 --> 00:17:55.260 Now this article vividly illustrates the reason for Friedman's stunning failure to recognize the housing and stock market bubbles that were very clearly evident and peaking by 2006. 179 00:17:55.260 --> 00:18:04.260 For Friedman did not give the slightest indication that he viewed the rapid growth in the money supply as a major factor driving the boom phases of the three episodes. 180 00:18:04.260 --> 00:18:11.260 Rather, he concluded, and I quote, monetary policy played a role in these three earlier booms, but only a supporting role. 181 00:18:11.260 --> 00:18:14.260 Technological change appears to be a major player. 182 00:18:14.260 --> 00:18:26.260 This conclusion perfectly exemplifies the mindset of price stabilizers like Friedman and his mentor Fisher, who he called the greatest American economist who ever lived. 183 00:18:26.260 --> 00:18:38.260 Since there was very little change in some arbitrarily selected price index in the US in the 1920s and 1990s, and in Japan in the 1980s, well then, monetary policy could not have been the cause of the boom. 184 00:18:38.260 --> 00:18:53.260 In the last section, so Friedman made no, his predictions of how the economy would continue under Greenspan, his pattern prediction, was devastatingly, blatantly, abysmally bad, okay, wrong. 185 00:18:53.260 --> 00:19:03.260 And then in the last section of my paper, I rehearse some of the predictions that were made by the Rothbardians, which Mark has written an article on, I sort of went over it. 186 00:19:03.260 --> 00:19:13.260 And I had also discovered that in 2004, in an article, I had sort of talked about the housing bubble and predicted that it would blow up, okay? 187 00:19:13.260 --> 00:19:23.260 And I'm just a guy who teaches at Pace University, I'm not Milton Friedman, so it's not me, it's just a good analytical economics that Austrian economics is 188 00:19:23.260 --> 00:19:32.260 that I think permits us to make these non-quantitative pattern predictions that Robbins and Mises talked about 189 00:19:32.260 --> 00:19:36.100 Mises talked about, and also Rothbard has a long section which I quote in my paper on 190 00:19:36.100 --> 00:19:40.700 pattern predictions. He does not use the term though. So I'll stop here and take any questions.