WEBVTT NOTE Does the Federal Debt Affect Corporate Profits, Employee Compensation, and Income Inequality? 1 00:00:00.000 --> 00:00:30.000 For those of you at the last session upstairs, I was a little shocked with the talk by someone from the Swiss Central Bank and he, the standard disclaimer, which is sort of routine if you've ever heard a talk by a Fed economist, they always start out and say, well, these are not my views of the Federal Bank or the organization I'm from, these are just my own views. I need to take that disclaimer one step further. I retired from the Kansas City Fed over a decade ago. If they thought I was trying to represent the Department 2 00:00:30.000 --> 00:00:53.000 So let me start out by trying to motivate my talk a little bit, and I'll just give you a few economic statistics here that I think will stand out and provide some motivation. 3 00:00:53.000 --> 00:00:59.000 This first one I'm sure won't surprise you too much, and that is at the end of the fiscal year 2011, 4 00:00:59.000 --> 00:01:07.600 In 2011, the federal debt was at almost $15 trillion, which is the highest since 1948. 5 00:01:07.600 --> 00:01:13.520 This is maybe not quite so widely known, which is that corporate profits is a share of national 6 00:01:13.520 --> 00:01:21.080 income or almost 15% of national income, and that's the highest since the Department of 7 00:01:21.080 --> 00:01:26.840 Commerce started collecting quarterly data way back in 1947. 8 00:01:26.840 --> 00:01:33.120 At the same time, these researchers here have pointed out that the labor share of national 9 00:01:33.120 --> 00:01:35.880 income is at a record low. 10 00:01:35.880 --> 00:01:40.560 So the question I'm addressing in this paper, is this all a coincidence? 11 00:01:40.560 --> 00:01:45.400 That the federal debt and corporate profits as a share of income are so high at the same 12 00:01:45.400 --> 00:01:47.720 time labor share is so low? 13 00:01:47.720 --> 00:01:51.360 And as you might expect, given that this is the way I introduced the talk, I'm going to 14 00:01:51.360 --> 00:01:53.280 claim it's not a coincidence. 15 00:01:53.280 --> 00:01:56.000 Let me apologize a little bit. 16 00:01:56.000 --> 00:01:59.280 So I've got overheads here so I see there's a little bit of glare here but I don't think 17 00:01:59.280 --> 00:02:05.880 there's anything I can do about it. So I'll just have to read over the parts where the 18 00:02:05.880 --> 00:02:12.800 glare is. So to put my talk in a broad perspective, there are two broad areas of research that 19 00:02:12.800 --> 00:02:16.860 I'm sort of bridging. One of them is people have looked at the effect of debt and deficits 20 00:02:16.860 --> 00:02:21.380 on the economy. And there's another branch of research that looked at how national income 21 00:02:21.380 --> 00:02:25.520 gets divided up between capital and labor. And to my knowledge there's no one that really 22 00:02:25.520 --> 00:02:30.120 is bridged these together, so that's kind of where my talk fits in and this paper fits 23 00:02:30.120 --> 00:02:33.040 in in this area of research. 24 00:02:33.040 --> 00:02:37.680 So to quickly summarize these two different areas to put my paper in context, I'll start 25 00:02:37.680 --> 00:02:40.440 out by just talking about what some people have talked about or found out when they've 26 00:02:40.440 --> 00:02:42.720 looked at the effect of debt and deficits. 27 00:02:42.720 --> 00:02:45.600 If any of you have worked in this area, let me apologize. 28 00:02:45.600 --> 00:02:52.920 I've taken thousands of pages of research and put them in nine bullet points so that 29 00:02:52.920 --> 00:03:22.920 Something that does get lost in the condensed version here, but if you're interested in learning more about this, the first two papers I've mentioned from Gail and Orsay Engen and Hubbard have excellent summaries of what's been going on, and of course part of the debate is how important are debt and deficits, and some people say well it's not really that big a deal, I'm sure there's going to be a lot of skeptics of that view here, but they've looked primarily at interest rate effects, and early research really did not find much of an effect of interest rates 30 00:03:22.920 --> 00:03:37.920 from Federal Debt and Deficits. That has changed with more recent research. There is now a consensus that raising the Federal debt or raising deficits do have an effect on Federal or the interest rate on Treasury securities. 31 00:03:37.920 --> 00:03:48.920 Of course, that's not the whole story. In addition to the effect on interest rates, of course, it's going to reduce national wealth. People do not boost their savings to compensate for large Federal deficits. 32 00:03:48.920 --> 00:03:55.920 There's not enough money coming in from abroad to compensate for federal deficits, so ultimately deficits will reduce national wealth. 33 00:03:55.920 --> 00:04:04.920 So that's a synopsis of that area of research. Let me do a quick synopsis of what people have decided when they've looked at the labor and capital share of income. 34 00:04:04.920 --> 00:04:12.920 And there's three different explanations as to why the profit share is so high and the labor share is so low. 35 00:04:12.920 --> 00:04:15.920 And the first one I've listed up there is the decline of unionization. 36 00:04:15.920 --> 00:04:20.300 But as we look at the private sector at least, there are many fewer people unionized today 37 00:04:20.300 --> 00:04:24.000 than there were 20, 30, 40 years ago. 38 00:04:24.000 --> 00:04:30.320 Another explanation that's given is globalization, and that is, with the opening up of Eastern 39 00:04:30.320 --> 00:04:36.180 Europe, the Chinese labor force, there's now a huge supply of low-skill labor. 40 00:04:36.180 --> 00:04:39.600 Unions can't compete without low-skill labor, or it's reduced their bargaining power, and 41 00:04:39.600 --> 00:04:42.700 therefore, labor share has declined. 42 00:04:42.700 --> 00:04:47.060 And finally, there's a technological explanation, which is that because technology is changing 43 00:04:47.060 --> 00:04:54.140 so fast, not just unionized workers, workers in general are finding their skills evaporating 44 00:04:54.140 --> 00:04:59.660 because of the technological change, because their skills are evaporating so fast, they're 45 00:04:59.660 --> 00:05:04.240 losing their bargaining power, and that's another reason labor share is declining. 46 00:05:04.240 --> 00:05:09.020 So if I've got to give you a consensus here, there is no consensus explanation, so that's 47 00:05:09.020 --> 00:05:13.980 That's where my paper is coming up with another alternative as to why, potentially, capital 48 00:05:13.980 --> 00:05:16.620 and labor shares are shifting. 49 00:05:16.620 --> 00:05:23.380 So to briefly describe the economics, and I apologize for some of the neoclassical terminology 50 00:05:23.380 --> 00:05:29.340 here, I heard the talk yesterday, Nikolai was sort of criticizing neoclassical terminology, 51 00:05:29.340 --> 00:05:33.540 but then he regressed later on and started using it, so I don't feel quite so defensive 52 00:05:33.540 --> 00:05:34.540 about using it. 53 00:05:34.540 --> 00:05:39.900 What happens when the government runs huge deficits is that it tracks from national savings. 54 00:05:39.900 --> 00:05:44.540 There's not enough resources for investment. So what happens is business investment goes 55 00:05:44.540 --> 00:05:49.460 down. You do not have as large a capital stock. With the lower capital stock, workers are 56 00:05:49.460 --> 00:05:55.060 less productive. The marginal product of labor goes down and wages go down. At the same time, 57 00:05:55.060 --> 00:06:01.020 with the lower capital stock, the value of that capital has gone up. That is, the marginal 58 00:06:01.020 --> 00:06:06.820 and the product of capital increases, and therefore businesses are earning higher profits. 59 00:06:06.820 --> 00:06:09.540 So again, as a final point to make here, and I'll come back to this maybe a little bit 60 00:06:09.540 --> 00:06:16.180 at the end, you know, one of the hot issues today in the public sphere is income inequality. 61 00:06:16.180 --> 00:06:21.460 The curiosity is this trend is going to actually make income inequality worse, the current 62 00:06:21.460 --> 00:06:23.060 increase in the federal debt. 63 00:06:23.060 --> 00:06:27.780 Okay, let's start looking at a little bit of the data here, and I won't go into this 64 00:06:27.780 --> 00:06:28.780 in too much detail. 65 00:06:28.780 --> 00:06:34.860 I'll show you one time chart here to get an idea of what the empirical foundation of this 66 00:06:34.860 --> 00:06:35.860 paper is. 67 00:06:35.860 --> 00:06:39.740 I'm looking at data from 1947 roughly to the current period. 68 00:06:39.740 --> 00:06:43.560 I actually end up dividing it up into business cycles. 69 00:06:43.560 --> 00:06:50.200 If you look at the trends I'm showing here, I'm showing profits first of all as a proportion 70 00:06:50.200 --> 00:06:55.820 of national income, and I'm showing business investment as a fraction of GDP. 71 00:06:55.820 --> 00:07:00.420 Unfortunately, these are about on the same scale, so I can plot them simultaneously looking 72 00:07:00.420 --> 00:07:03.460 at the left-hand side here. 73 00:07:03.460 --> 00:07:07.900 You can see these things go up and down over time, and there are business cycle effects, 74 00:07:07.900 --> 00:07:12.220 like right here in the early 90s, both of these moved together, they moved together 75 00:07:12.220 --> 00:07:16.700 here in the late 70s, early 80s, and so on, so you can see co-movement above business 76 00:07:16.700 --> 00:07:17.700 cycles. 77 00:07:17.700 --> 00:07:24.100 However, here in the late 70s when business investment was very high, the profits were 78 00:07:24.100 --> 00:07:25.940 are relatively low. 79 00:07:25.940 --> 00:07:30.540 So I want to separate the business cycle effects from kind of the longer term trends. 80 00:07:30.540 --> 00:07:37.620 So my preferred way of looking at this data is as follows, here again the same data, its 81 00:07:37.620 --> 00:07:42.960 profits as a share of income, its investment as a share of GDP, and I've divided it up 82 00:07:42.960 --> 00:07:48.500 into 10 business cycles, there are 10 full business cycles since 1947, and I've plotted 83 00:07:48.500 --> 00:07:53.320 them on this scatter chart, where I've got profits as a share of income here, business 84 00:07:53.320 --> 00:08:10.320 This investment is the share of GDP here, and you can see when investment is very high here, the profits are low and vice versa. There's an inverse relationship there. The trend line is negatively sloped. 85 00:08:10.320 --> 00:08:21.320 So for those of you involved in corporate finance, this may be a little bit surprising because businesses tend to take a lot of their funding from finance, from internal finance, from retained earnings. 86 00:08:21.320 --> 00:08:24.760 So that would suggest that when profits are high, you'd have a lot of retained earnings 87 00:08:24.760 --> 00:08:28.920 and a high investment, and that's certainly true through the business cycle, but this 88 00:08:28.920 --> 00:08:33.340 chart suggests that that is not true over the longer term, that what really dominates 89 00:08:33.340 --> 00:08:37.760 is when you have too much investment, profits actually go down. 90 00:08:37.760 --> 00:08:42.000 So that chart links profits and investment. 91 00:08:42.000 --> 00:08:47.040 Of course, ultimately I want to talk about the debt, so let me see how debt behaves here. 92 00:08:47.040 --> 00:08:50.640 So this shows how debt varies, or how investment varies with debt. 93 00:08:50.640 --> 00:08:56.840 So again, here on the horizontal axis, I've got debt as a fraction of GDP. On the vertical 94 00:08:56.840 --> 00:09:01.200 axis, I have business investment as a GDP. And again, if you believe in crowding out, 95 00:09:01.200 --> 00:09:06.480 and this is really a crowding out story, federal debt, crowd out business investment. When 96 00:09:06.480 --> 00:09:11.040 the debt ratio is high, investment is low and vice versa. 97 00:09:11.040 --> 00:09:18.360 Okay, so now I've linked profits and investment. I've linked investment and debt. So let's 98 00:09:18.360 --> 00:09:23.520 Let's just sort of look right across the chasm here and look, so what's the relationship 99 00:09:23.520 --> 00:09:31.560 between profits and debt? And of course, from those other two, this is exactly what you'd 100 00:09:31.560 --> 00:09:36.620 expect when debt gets high and crowds things out, profits go up, so we've got a positive 101 00:09:36.620 --> 00:09:42.120 relationship between debt and profits. Again, looking at business cycle averages, each one 102 00:09:42.120 --> 00:09:47.920 of these data points is an average of the ten post-war business cycles. And the other 103 00:09:47.920 --> 00:09:53.560 I saw that, of course, if profits are going up when it goes up, then employee compensation 104 00:09:53.560 --> 00:10:00.680 or the labor share of income should go down, and again, that shows up in that chart. 105 00:10:00.680 --> 00:10:03.920 So to me, I thought that'd be sort of the most interesting part of the empirical results 106 00:10:03.920 --> 00:10:08.520 of my paper, but I had a special request over lunch that someone insisted on seeing regressions 107 00:10:08.520 --> 00:10:15.480 from the Fed economists, so I've got some regressions of the same information. 108 00:10:15.480 --> 00:10:18.940 So my dependent variables here, again, I'm mostly concerned about with the profit share 109 00:10:18.940 --> 00:10:25.880 and the labor share of income. So those are over here. Here I've got some independent 110 00:10:25.880 --> 00:10:29.320 variables. I don't have time to sort of talk about how I chose these. For those of you 111 00:10:29.320 --> 00:10:32.760 who are econometricians, it's largely ad hoc, even though I'm copying what other people 112 00:10:32.760 --> 00:10:37.120 did that was ad hoc. So I would claim this is less ad hoc than what they did. But the 113 00:10:37.120 --> 00:10:44.440 main point is, again, the debt to GDP ratio is significant in these regressions. Negative 114 00:10:44.440 --> 00:10:49.480 Positive in the case of compensation, positive in the case of profits. 115 00:10:49.480 --> 00:10:51.240 So that's pretty much the story. 116 00:10:51.240 --> 00:10:53.120 Where do I go from here? 117 00:10:53.120 --> 00:10:56.320 Potentially looking at international data might be interesting. 118 00:10:56.320 --> 00:11:02.320 It turns out the same things have been observed internationally, at least in other industrialized 119 00:11:02.320 --> 00:11:03.320 economies. 120 00:11:03.320 --> 00:11:07.800 Profit shares have gone up, labor shares have gone down, and of course debt has also gone 121 00:11:07.800 --> 00:11:08.800 up. 122 00:11:08.800 --> 00:11:13.600 Curiously, if you look at developing countries, they tend to also have lower labor shares 123 00:11:13.600 --> 00:11:19.320 and Profit Shares, so to the extent that if we spend all our national wealth through the 124 00:11:19.320 --> 00:11:24.040 budget deficit, we're going to start looking more like developing countries, so this is 125 00:11:24.040 --> 00:11:28.600 consistent with the hypothesis in my paper. 126 00:11:28.600 --> 00:11:33.200 So to finally wrap this up, let me just say that I think this is a good example of the 127 00:11:33.200 --> 00:11:38.720 law of unintended consequences, and that is people sometimes justify spending a lot of 128 00:11:38.720 --> 00:11:43.500 of Federal Money because we need to combat this evil thing of income inequality and the 129 00:11:43.500 --> 00:11:48.460 unintended consequences, high debt is driving up the profit share, driving down the labor 130 00:11:48.460 --> 00:11:54.460 share and having exactly the opposite consequence than was intended. Thank you.