WEBVTT NOTE Did Increased Required Reserves Prolong the Great Depression? 1 00:00:00.000 --> 00:00:06.000 The paper that I am presenting is actually co-authored with my colleague Jeff Herbner. 2 00:00:06.000 --> 00:00:12.000 The paper entitled, Did Increased Reserve, Required Reserves Prolong the Great Depression? 3 00:00:12.000 --> 00:00:16.000 The short answer is no. At least that's what we're going to argue. 4 00:00:16.000 --> 00:00:22.000 And the reason we're interested in this question is simply because of really monetary policy, 5 00:00:22.000 --> 00:00:27.000 more recent monetary policy during what we could call the Great Recession. 6 00:00:27.000 --> 00:00:35.760 Since the beginning of 2008, the monetary base has increased by 178% because the Chairman 7 00:00:35.760 --> 00:00:40.800 of the Fed wants it that way, to easily the highest level ever. 8 00:00:40.800 --> 00:00:47.440 And since the beginning of 2008, excess reserves in the banking system in commercial banks 9 00:00:47.440 --> 00:00:55.640 increased in the United States, increased from $1.65 billion to $1.037 trillion with 10 00:00:55.640 --> 00:01:22.640 If banks decide to expand credit in the form of demand deposits while maintaining, if they decide now we're finally actually going to start extending credit, seriously extending credit, while maintaining simply a 20% reserve ratio, that would allow M2 to increase by $5.185 trillion. 11 00:01:22.640 --> 00:01:36.640 If banks do the same while maintaining a more conventional 10% reserve ratio, M2 would increase by $10.37 trillion, and this would more than double M2. 12 00:01:36.640 --> 00:01:46.640 So, naturally, people have considered the question, what should we do besides run for our lives? 13 00:01:46.640 --> 00:01:52.640 And Ben Bernanke has told us and assured us we have a plan. 14 00:01:52.640 --> 00:01:53.640 We have a plan. 15 00:01:53.640 --> 00:02:01.640 Part of that plan included paying commercial banks interest on their reserve deposits to keep them in the system. 16 00:02:01.640 --> 00:02:07.640 Other people have pointed out that that doesn't really unwind the reserves, that just keeps them in the system. 17 00:02:07.640 --> 00:02:11.640 So what do we do? Well, we have a plan. 18 00:02:11.640 --> 00:02:24.640 One of the things that we could do to try to mitigate the potential serious inflationary impact of these increased excess reserves 19 00:02:24.640 --> 00:02:27.640 would be raise legal reserve requirements. 20 00:02:27.640 --> 00:02:35.640 This would constrain how much of the additional monetary base could actually be extended as credit 21 00:02:35.640 --> 00:02:41.880 and hence constrain the amount of new demand deposits and new money that is actually created 22 00:02:41.880 --> 00:02:44.000 by the banking system. 23 00:02:44.000 --> 00:02:48.240 And this is an idea that is out there. 24 00:02:48.240 --> 00:02:53.880 A number of people, a number of economists and policy makers, however, aren't too excited 25 00:02:53.880 --> 00:02:55.400 about this. 26 00:02:55.400 --> 00:03:01.860 David C. Wheelock, vice president of the St. Louis Fed in a little newsletter called Economic 27 00:03:01.860 --> 00:03:14.140 In 2009, he said that referring to the experience of the 1930s, he said experience demonstrates 28 00:03:14.140 --> 00:03:18.500 that raising reserve requirements is surely not the best way to eliminate excess reserves. 29 00:03:18.500 --> 00:03:20.020 It's not the best way. 30 00:03:20.020 --> 00:03:24.020 You could eliminate them this way, but it's not the best way because of negative economic 31 00:03:24.020 --> 00:03:26.420 consequences. 32 00:03:26.420 --> 00:03:36.420 Christina Romer, in an article in The Economist in June 18, 2009 entitled The Lessons of 1937, said this quote, 33 00:03:36.420 --> 00:03:41.420 In 1936, the Federal Reserve began to worry about its exit strategy. 34 00:03:41.420 --> 00:03:50.420 After several years of loose monetary policy, American banks were withholding large quantities of reserves in excess of their legislative requirements. 35 00:03:50.420 --> 00:03:58.420 Monetary policymakers feared the excess reserves would make it difficult to tighten if inflation developed. 36 00:03:58.420 --> 00:04:03.420 The Fed then doubled reserve requirements in a series of steps. It was a three-step process. 37 00:04:03.420 --> 00:04:10.420 And when the excess was legislated away, they scrambled to replace it by reducing lending. 38 00:04:10.420 --> 00:04:19.420 This contracted the money supply, and that ushered in then the recession and crash in 1938, which prolonged the Great Depression. 39 00:04:19.420 --> 00:04:27.920 and she explicitly cites the empirical work done by Friedman and Schwartz to demonstrate this case. 40 00:04:27.920 --> 00:04:40.420 And so people like Wheelock and Romer and Paul Casreal at Northern Trust follow Friedman and Schwartz 41 00:04:40.420 --> 00:04:48.920 in their warning us of the problems of raising excess reserve requirements to sop up excess reserves 42 00:04:48.920 --> 00:04:51.920 and make them non-excess. 43 00:04:51.920 --> 00:04:54.120 And so we wanted to investigate this hypothesis. 44 00:04:54.120 --> 00:04:55.760 Is this really the case? 45 00:04:55.760 --> 00:04:57.680 Is the Friedman-Schwarz hypothesis, 46 00:04:57.680 --> 00:05:02.120 is it a sound explanation for why the Great Depression 47 00:05:02.120 --> 00:05:05.080 lingered past 37? 48 00:05:05.080 --> 00:05:06.960 And Friedman and Schwarz's thesis 49 00:05:06.960 --> 00:05:09.600 is essentially that the three-step increase 50 00:05:09.600 --> 00:05:14.760 in legal reserve requirements during 1936 and 37 51 00:05:14.760 --> 00:05:16.960 had deflationary effects. 52 00:05:16.960 --> 00:05:22.120 Because although commercial bank reserves were in excess in a legal sense, banks did 53 00:05:22.120 --> 00:05:28.360 not perceive them as excess in a prudential or economic sense. 54 00:05:28.360 --> 00:05:33.680 Increasing required reserves debased, or I'm sorry, decreased the current quantity of reserves 55 00:05:33.680 --> 00:05:38.200 banks could draw upon to meet depositors' demand for currency. 56 00:05:38.200 --> 00:05:43.440 Banks therefore, it is argued, took steps to increase their reserves even further. 57 00:05:43.440 --> 00:05:53.000 And this effort to further increase bank reserves resulted in a contraction of the money supply. 58 00:05:53.000 --> 00:05:57.600 And despite commercial banks holding excess legal reserves then, the increase in legal 59 00:05:57.600 --> 00:06:03.480 reserve requirements did have real economic effects at the time of the increase and not 60 00:06:03.480 --> 00:06:10.680 just in the future by restraining future excess credit expansion, which was really the desire 61 00:06:10.680 --> 00:06:16.340 of the Federal Reserve governors at the time. 62 00:06:16.340 --> 00:06:21.680 As Friedman and Schwartz tell their story of monetary history leading up to the recession 63 00:06:21.680 --> 00:06:27.400 in the Depression of 38, they argue that in 1936, banks began to accumulate excess reserves 64 00:06:27.400 --> 00:06:30.640 because they thought it was the prudential thing to do. 65 00:06:30.640 --> 00:06:35.640 Commercial banks, they argue, lost faith in the Federal Reserve as a lender of last resort 66 00:06:35.640 --> 00:06:45.440 and recognize that holding only legally reserved, legally required reserves was not enough to ward off insolvency in times of crisis. 67 00:06:45.440 --> 00:06:51.240 Consequently, banks accumulated cash assets because their demand for liquidity had arisen. 68 00:06:51.240 --> 00:07:00.540 And when the Federal Reserve raised Federal Reserve requirements in 1936 and 37, this immobilized that cash for use. 69 00:07:00.540 --> 00:07:06.240 So banks therefore built up even more reserves and increased their ratio of cash to total assets. 70 00:07:06.240 --> 00:07:14.500 The building up of reserves proved to be deflationary, they argue, causing the depression of 1938. 71 00:07:14.500 --> 00:07:22.780 Now a survey of the relevant economic literature reveals many who accept Friedman and Schwartz. 72 00:07:22.780 --> 00:07:29.200 Richard Timberlake, for example, argues in his work that was published primarily in the Friedman, 73 00:07:29.200 --> 00:07:34.360 as well as in his book on the history of monetary policy in the U.S. 74 00:07:34.360 --> 00:07:39.360 Timberlake argues that the Federal Reserve needlessly, as he puts it, and foolishly was 75 00:07:39.360 --> 00:07:44.800 concerned about inflation as excess reserves piled up in the mid-1930s. 76 00:07:44.800 --> 00:07:47.880 And I've heard similar language today. 77 00:07:47.880 --> 00:07:50.040 Inflation, where's the inflation? 78 00:07:50.040 --> 00:07:55.600 It's foolish to look around and worry about inflation now when prices are staying relatively 79 00:07:55.600 --> 00:08:19.600 Timberlake said that the policy of the Fed in the 30s is also foolish. He also agrees explicitly with Friedman and Schwartz that the U.S. Treasury reinforced the Federal Reserve's deflationary policies by sterilizing the influx of gold. 80 00:08:19.600 --> 00:08:26.600 We don't really deal with that element in much detail yet in this paper, which is still a paper in progress. 81 00:08:26.600 --> 00:08:32.600 But the heart of Timberlake's argument is to construct a measure of potential money supply 82 00:08:32.600 --> 00:08:36.600 and full employment money supply for the time period in question. 83 00:08:36.600 --> 00:08:40.600 And on the basis of what he perceives to be the full employment money supply, 84 00:08:40.600 --> 00:08:45.600 in other words the money supply that we should have had to maintain full employment in 36 and 37, 85 00:08:45.600 --> 00:08:52.600 He argues that the increases in the reserve requirements kept the money supply much below its full employment level. 86 00:08:52.600 --> 00:08:58.600 And so on the basis of this analysis, he repeats the conclusion of Friedman and Schwartz, almost word for word. 87 00:08:58.600 --> 00:09:06.600 Although commercial banks held large quantity of reserves that were excess legally, they were not economically in excess because the banks needed them 88 00:09:06.600 --> 00:09:11.600 because they couldn't trust the Federal Reserve to be a provider of adequate liquidity. 89 00:09:11.600 --> 00:09:20.600 Now, we note in our paper that Tiver Lake's definition of the full employment money supply level is not without its own problems. 90 00:09:20.600 --> 00:09:27.600 For instance, in identifying the supply that would yield full employment, he makes two adjustments to the existing money stock, 91 00:09:27.600 --> 00:09:36.600 and this is what he says, quote, one, the first adjustment is the additional amount that would have been necessary to accommodate the real output of goods 92 00:09:36.600 --> 00:09:43.240 of Goods and Services to a Fully-Employed Labor Force would, let me read that again. 93 00:09:43.240 --> 00:09:47.960 One, the additional amount that would have been necessary to accommodate the real output 94 00:09:47.960 --> 00:09:51.680 of goods and services a fully-employed labor force would produce. 95 00:09:51.680 --> 00:09:55.840 And two, this is the second adjustment, the additional amount of money that would have 96 00:09:55.840 --> 00:10:01.720 been necessary to get the price level back up to the value it had had in 1929 before 97 00:10:01.720 --> 00:10:06.040 the current cyclical decline began, end of quote. 98 00:10:06.040 --> 00:10:12.920 Now note that Timberlake here uses June 1929 prices as a benchmark as if conditions in 99 00:10:12.920 --> 00:10:19.280 June 1929 were normal, not the height of an inflationary boom, which I find personally 100 00:10:19.280 --> 00:10:22.040 somewhat problematic. 101 00:10:22.040 --> 00:10:28.280 Other economists besides Timberlake has agreed with Friedman and Schwartz, Meltzer in his 102 00:10:28.280 --> 00:10:42.280 This history of the Fed essentially adopts Friedman and Schwartz's story as one, the story that increases the reserves for deflationary is one contributing factor to prolonging the Great Depression. 103 00:10:42.280 --> 00:10:54.280 Christina Romer in a paper on what ended the Great Depression that she published in 1992 approvingly cites Friedman and Schwartz in her paper, just sort of accepts it as the case. 104 00:10:54.280 --> 00:11:00.280 However, there's not monolithic support for Friedman and Schwartz in the economics literature. 105 00:11:00.280 --> 00:11:11.280 Over ten years before Friedman and Schwartz wrote their work, Benjamin Anderson criticized in detail the hypothesis that raising reserve requirements lengthened the Great Depression. 106 00:11:11.280 --> 00:11:19.280 Anderson staunchly maintained that the crisis of late 37 and 1938 were in no way caused by the increase in reserve requirements. 107 00:11:19.280 --> 00:11:27.280 In supporting his argument, he noted that the increased reserve requirements coincided with very little changes in commercial interest rates. 108 00:11:27.280 --> 00:11:31.280 In fact, most interest rates fell during that period of time. 109 00:11:31.280 --> 00:11:41.280 Some rates increased slightly, some fell slightly during the relevant time period, most was in some level of decline during the relevant time period. 110 00:11:41.280 --> 00:11:46.120 and at the same time commercial loans increased from approximately 4.8 billion dollars in 111 00:11:46.120 --> 00:11:53.240 June 35 to almost 7 billion by the end of 1937. 112 00:11:53.240 --> 00:11:58.200 In summary Anderson concludes that no one who was credit worthy was denied credit due 113 00:11:58.200 --> 00:12:02.120 to the increased reserve requirements and consequently they did not interfere with commercial 114 00:12:02.120 --> 00:12:06.980 activity, commercial bank deposits, issuing of securities or any increase in the price 115 00:12:06.980 --> 00:12:11.380 of Securities. 116 00:12:11.380 --> 00:12:17.820 Timberlake's analysis of the Federal Reserve requirements, or Federal Reserve's required 117 00:12:17.820 --> 00:12:23.980 reserve policy was criticized explicitly by Joe Salerno. 118 00:12:23.980 --> 00:12:29.460 Using Timberlake's own statistics, Salerno points out that from June 36 to June 1937, 119 00:12:29.460 --> 00:12:33.700 the money supply continued to grow, didn't fall at all. 120 00:12:33.700 --> 00:12:44.100 And so what we then do is take a look at the data, monetary data and interest rate data 121 00:12:44.100 --> 00:12:53.120 and lending data, and look to see whether the sort of post-Hawk, ergo propter hawk plausibility 122 00:12:53.120 --> 00:12:57.620 of the Friedman-Schwarz hypothesis actually holds water. 123 00:12:57.620 --> 00:13:03.620 And looking at the data, we note that the first problem with Friedman-Schwarz's 124 00:13:03.620 --> 00:13:10.420 This explanation is that there was no great increase of bank reserves from August 36 to 125 00:13:10.420 --> 00:13:15.660 December 37, as Friedman and Schwartz themselves indicate. 126 00:13:15.660 --> 00:13:26.020 There was some increase in reserves, but not a great increase, not one that you would think 127 00:13:26.020 --> 00:13:32.180 would be large enough to cause a great monetary contraction. 128 00:13:32.180 --> 00:13:42.780 Additionally, if we look, for instance, from the start of the period of increasing reserve ratios in August 36 until January 1938, 129 00:13:42.780 --> 00:13:51.180 excess reserves continue to fall, reaching their lowest point by the end of December 1937. 130 00:13:51.180 --> 00:13:58.180 Total loans also fell from June 37 to December 37, 131 00:13:58.180 --> 00:14:06.680 But that cannot be due to a build-up of excess reserves, because they did not build up excess reserves during that time period. 132 00:14:06.680 --> 00:14:12.180 The decline, we argue, is most likely from reduced demand for loans. 133 00:14:12.180 --> 00:14:24.180 On the demand side, it was only in the first half of 1938 that total loans declined at the same time that excess reserves rose. 134 00:14:24.180 --> 00:14:35.680 The second problem that we find with Friedman's Theory is that Friedman's Theory requires a policy decline in the money stock, not just bank loans themselves. 135 00:14:35.680 --> 00:14:43.180 But the contraction in M1 in 1937 was caused entirely, we argue, from a decline in demand deposits. 136 00:14:43.180 --> 00:14:51.680 Since the banks were not building excess reserves, it's likely that this effect, too, was caused by reduced demand for these deposits. 137 00:14:51.680 --> 00:14:59.680 People were pulling out bank deposits and moving into other things like currency and non-bank assets. 138 00:14:59.680 --> 00:15:11.680 The contraction of M2, which is Friedman's preferred monetary aggregate in 1937, was even smaller than that of M1 because time deposits continued to increase during this period. 139 00:15:11.680 --> 00:15:40.680 As I also already mentioned, interest rates generally fell broadly throughout this period of time, which provides evidence that a reduced supply of credit was not the driving factor behind a reduction in loans, but a drop in the demand for loanable funds. 140 00:15:40.680 --> 00:15:55.680 If there was a drop in the quantity of loans due to a decrease in credit being made available by banks, we would expect a significant increase in interest rates, and this was not the case. 141 00:15:58.680 --> 00:16:09.680 So we argue, as is clear from the theoretical and empirical analysis, the preponderance of evidence indicates that the Federal Reserve's raising of the legal reserve requirements were not the cause of the recession of 1938. 142 00:16:09.680 --> 00:16:13.680 and one possible implication is that in the presence of the extremely large 143 00:16:13.680 --> 00:16:17.680 quantity of excess reserves in our contemporary banking system, raising 144 00:16:17.680 --> 00:16:21.680 legal requirements should be considered as an effective or 145 00:16:21.680 --> 00:16:25.680 relatively effective method of unwinding 146 00:16:25.680 --> 00:16:29.680 our excess reserves with relatively little 147 00:16:29.680 --> 00:16:33.680 monetary destabilization. Thank you.