WEBVTT 00:00:00.000 --> 00:00:03.444 align:middle line:84% [SQUEAKING] [RUSTLING] [CLICKING] 00:00:03.444 --> 00:00:16.328 align:middle line:90% 00:00:16.328 --> 00:00:17.870 align:middle line:84% RICARDO CABALLERO: So today I'm going 00:00:17.870 --> 00:00:22.040 align:middle line:84% to talk about the Phillips curve and inflation. 00:00:22.040 --> 00:00:25.070 align:middle line:84% Now as I said in the previous lecture, 00:00:25.070 --> 00:00:27.680 align:middle line:84% the material that is specific to this lecture 00:00:27.680 --> 00:00:29.750 align:middle line:90% will not enter this quiz. 00:00:29.750 --> 00:00:32.960 align:middle line:84% It's the beginning of what is perhaps the most important model 00:00:32.960 --> 00:00:38.270 align:middle line:84% you'll see in this class, but it will take us 00:00:38.270 --> 00:00:40.580 align:middle line:84% three or four lectures to develop. 00:00:40.580 --> 00:00:45.080 align:middle line:84% So I'm going to say things that certainly will-- 00:00:45.080 --> 00:00:46.970 align:middle line:84% may help you understand a little better 00:00:46.970 --> 00:00:50.510 align:middle line:84% the previous lecture, so if you're only 00:00:50.510 --> 00:00:53.420 align:middle line:84% concerned about the next quiz, there 00:00:53.420 --> 00:00:57.590 align:middle line:84% will be a sort of small review of the previous lecture here. 00:00:57.590 --> 00:01:01.040 align:middle line:84% But again, anything that is specific to this lecture 00:01:01.040 --> 00:01:06.510 align:middle line:84% and was not in the previous one won't be part of this quiz. 00:01:06.510 --> 00:01:08.120 align:middle line:90% So what is this Phillips curve? 00:01:08.120 --> 00:01:15.980 align:middle line:84% Well, in 1958, an economist at LSE, London School of Economics, 00:01:15.980 --> 00:01:18.720 align:middle line:84% came out with just an empirical relationship. 00:01:18.720 --> 00:01:24.840 align:middle line:84% This is AW Phillips found that using historical data 00:01:24.840 --> 00:01:30.690 align:middle line:84% for the US, I think he did, there was a negative relation up 00:01:30.690 --> 00:01:33.480 align:middle line:90% to the '50s, I think. 00:01:33.480 --> 00:01:35.340 align:middle line:84% There was a negative relationship 00:01:35.340 --> 00:01:39.600 align:middle line:84% between the unemployment rate and the rate of inflation. 00:01:39.600 --> 00:01:43.980 align:middle line:84% And then our very own Paul Samuelson and Robert Solow 00:01:43.980 --> 00:01:46.680 align:middle line:84% labeled this relationship the Phillips curve 00:01:46.680 --> 00:01:49.560 align:middle line:90% in honor of AW Phillips. 00:01:49.560 --> 00:01:55.890 align:middle line:84% And nowadays, it is a central concept in macroeconomics, 00:01:55.890 --> 00:01:58.920 align:middle line:84% and it's certainly very, very relevant 00:01:58.920 --> 00:02:01.980 align:middle line:84% to understand what is going on right now, 00:02:01.980 --> 00:02:03.870 align:middle line:84% not only in the US economy, but in most 00:02:03.870 --> 00:02:06.070 align:middle line:90% economies around the world. 00:02:06.070 --> 00:02:12.120 align:middle line:84% So let me show you this is not the one that Phillips plotted. 00:02:12.120 --> 00:02:17.340 align:middle line:84% I think this is the one that Samuelson and Solow plotted. 00:02:17.340 --> 00:02:23.040 align:middle line:84% For data from between 1900 and 1960 for the US, 00:02:23.040 --> 00:02:25.980 align:middle line:84% you find this negative correlation. 00:02:25.980 --> 00:02:30.690 align:middle line:84% I think it's reasonable, this negative correlation 00:02:30.690 --> 00:02:35.220 align:middle line:84% between the unemployment rate and inflation rate, no? 00:02:35.220 --> 00:02:36.870 align:middle line:84% At very low levels of unemployment, 00:02:36.870 --> 00:02:39.660 align:middle line:84% you typically see very high levels of inflation. 00:02:39.660 --> 00:02:42.630 align:middle line:84% Conversely, very high levels of unemployment, 00:02:42.630 --> 00:02:46.680 align:middle line:84% you tend to see low levels of inflation or even deflation. 00:02:46.680 --> 00:02:49.920 align:middle line:84% In fact, this period includes the Great Depression, 00:02:49.920 --> 00:02:52.180 align:middle line:90% for example. 00:02:52.180 --> 00:02:55.620 align:middle line:90% So that's the data. 00:02:55.620 --> 00:02:58.820 align:middle line:84% And again, this was just an empirical regularity. 00:02:58.820 --> 00:03:02.570 align:middle line:84% But we can build some theory about this relationship using 00:03:02.570 --> 00:03:05.590 align:middle line:84% the ingredients-- most of the ingredients that-- 00:03:05.590 --> 00:03:08.390 align:middle line:84% I mean, essentially we can build a relationship 00:03:08.390 --> 00:03:12.830 align:middle line:84% that is downward sloping from the ingredients we already have. 00:03:12.830 --> 00:03:15.440 align:middle line:84% And this is the part that is a little bit of a review 00:03:15.440 --> 00:03:17.060 align:middle line:90% of the previous lecture. 00:03:17.060 --> 00:03:20.930 align:middle line:90% Remember that we had a-- 00:03:20.930 --> 00:03:22.490 align:middle line:84% actually, the previous two lectures. 00:03:22.490 --> 00:03:27.980 align:middle line:84% We had a wage setting equation, W equal expected prices, 00:03:27.980 --> 00:03:30.290 align:middle line:84% and then a decreasing function of unemployment 00:03:30.290 --> 00:03:34.310 align:middle line:84% and an increasing function of this labor market supporting 00:03:34.310 --> 00:03:38.390 align:middle line:84% institutions, or workers supporting institutions. 00:03:38.390 --> 00:03:41.770 align:middle line:84% Institutional variables, I should say. 00:03:41.770 --> 00:03:43.850 align:middle line:84% And then we had a price setting equation, which 00:03:43.850 --> 00:03:48.050 align:middle line:90% was simply the wage marked up. 00:03:48.050 --> 00:03:50.010 align:middle line:90% m is a positive constant. 00:03:50.010 --> 00:03:52.730 align:middle line:90% So let me start from these two. 00:03:52.730 --> 00:03:56.002 align:middle line:84% So what I'm trying to do is derive a Phillips curve. 00:03:56.002 --> 00:03:57.960 align:middle line:84% Again, this was only an empirical relationship. 00:03:57.960 --> 00:04:00.390 align:middle line:84% But it turns out that even with theory, we 00:04:00.390 --> 00:04:03.870 align:middle line:84% knew by the time of Samuelson and Solow 00:04:03.870 --> 00:04:07.380 align:middle line:84% we could come up with a theory of that relationship. 00:04:07.380 --> 00:04:09.300 align:middle line:84% And that theory builds on the ingredients 00:04:09.300 --> 00:04:10.570 align:middle line:90% we have been looking at. 00:04:10.570 --> 00:04:12.750 align:middle line:84% So these are the wage setting equation, 00:04:12.750 --> 00:04:14.170 align:middle line:90% the price setting equation. 00:04:14.170 --> 00:04:16.560 align:middle line:84% I'm going to just simplify things and assume 00:04:16.560 --> 00:04:20.459 align:middle line:84% that this relationship here, this function, F of u z, 00:04:20.459 --> 00:04:22.920 align:middle line:84% is some linear function, at least locally 00:04:22.920 --> 00:04:26.550 align:middle line:84% linear function, which is decreasing in unemployment 00:04:26.550 --> 00:04:28.590 align:middle line:90% and increasing in z. 00:04:28.590 --> 00:04:30.570 align:middle line:84% Why is it decreasing in unemployment? 00:04:30.570 --> 00:04:35.260 align:middle line:90% 00:04:35.260 --> 00:04:38.800 align:middle line:84% This says no, that if unemployment goes up 00:04:38.800 --> 00:04:44.050 align:middle line:84% for any given expected price, wage demand is lower. 00:04:44.050 --> 00:04:49.420 align:middle line:84% And that's essentially because for the worker, it's sort of-- 00:04:49.420 --> 00:04:52.120 align:middle line:84% becoming unemployed is a scary situation. 00:04:52.120 --> 00:04:59.920 align:middle line:84% Conversely for firms, it's easier to find a worker and-- 00:04:59.920 --> 00:05:02.450 align:middle line:84% so we say a worker is scared for two reasons. 00:05:02.450 --> 00:05:05.560 align:middle line:84% One is that it's more likely that he gets fired 00:05:05.560 --> 00:05:07.150 align:middle line:90% when unemployment is high. 00:05:07.150 --> 00:05:08.800 align:middle line:90% Typically, that's a recession. 00:05:08.800 --> 00:05:12.130 align:middle line:84% It's also like-- that worker knows 00:05:12.130 --> 00:05:15.670 align:middle line:84% that if she were to fall into the unemployment pool, 00:05:15.670 --> 00:05:18.460 align:middle line:84% it will take a longer time to get out of it. 00:05:18.460 --> 00:05:20.500 align:middle line:84% And the firms are seeing the opposite side. 00:05:20.500 --> 00:05:22.720 align:middle line:84% It's pretty easy for them to replace a worker 00:05:22.720 --> 00:05:24.880 align:middle line:84% if they were to dismiss a worker, 00:05:24.880 --> 00:05:27.670 align:middle line:84% because there's lots of available workers 00:05:27.670 --> 00:05:28.970 align:middle line:90% in unemployment. 00:05:28.970 --> 00:05:30.560 align:middle line:84% So that's the reason that's negative. 00:05:30.560 --> 00:05:34.580 align:middle line:84% So I'm going to stick this function back in here. 00:05:34.580 --> 00:05:38.300 align:middle line:84% And then I'm going to replace this W with this function 00:05:38.300 --> 00:05:41.000 align:middle line:84% in there in the price setting equation, 00:05:41.000 --> 00:05:43.790 align:middle line:84% and I end up with an equation for P. 00:05:43.790 --> 00:05:50.180 align:middle line:84% So this says that the price, given the expected price, 00:05:50.180 --> 00:05:53.450 align:middle line:84% is decreasing unemployment, and increasing in z, 00:05:53.450 --> 00:05:55.920 align:middle line:90% and increasing in the markup. 00:05:55.920 --> 00:05:59.960 align:middle line:84% So again, why is this price decreasing in unemployment? 00:05:59.960 --> 00:06:02.600 align:middle line:90% 00:06:02.600 --> 00:06:05.630 align:middle line:84% This is the part that is review of the previous lecture. 00:06:05.630 --> 00:06:06.603 align:middle line:90% Previous two lectures. 00:06:06.603 --> 00:06:08.130 align:middle line:84% STUDENT: Because when wages go down, 00:06:08.130 --> 00:06:10.108 align:middle line:84% and then factors of production are achieved. 00:06:10.108 --> 00:06:11.400 align:middle line:90% RICARDO CABALLERO: OK, perfect. 00:06:11.400 --> 00:06:14.820 align:middle line:84% Because wages go down, since our firm 00:06:14.820 --> 00:06:16.980 align:middle line:84% needs one worker to produce one unit of a good, 00:06:16.980 --> 00:06:20.550 align:middle line:84% then the cost of production of one unit goes down with the wage 00:06:20.550 --> 00:06:23.370 align:middle line:84% and therefore the price goes down. 00:06:23.370 --> 00:06:26.910 align:middle line:84% Because the firm is asking for a constant markup over that wage, 00:06:26.910 --> 00:06:31.310 align:middle line:84% the wage declines, and the price drops. 00:06:31.310 --> 00:06:31.810 align:middle line:90% Good. 00:06:31.810 --> 00:06:33.770 align:middle line:90% So that's all review. 00:06:33.770 --> 00:06:36.790 align:middle line:84% So this equation, you have seen just 00:06:36.790 --> 00:06:39.520 align:middle line:84% without an explicit functional form here. 00:06:39.520 --> 00:06:41.798 align:middle line:84% What I want to do is to go from here-- 00:06:41.798 --> 00:06:43.340 align:middle line:84% this is still not the Phillips curve. 00:06:43.340 --> 00:06:46.510 align:middle line:84% Remember, the Phillips curve was a relationship between inflation 00:06:46.510 --> 00:06:47.890 align:middle line:90% and unemployment. 00:06:47.890 --> 00:06:51.040 align:middle line:84% Here, we have a relationship between the price level 00:06:51.040 --> 00:06:52.750 align:middle line:90% and unemployment. 00:06:52.750 --> 00:06:56.990 align:middle line:84% So we want to take one derivative higher. 00:06:56.990 --> 00:06:59.452 align:middle line:84% We want to go to relationship between inflation 00:06:59.452 --> 00:07:00.160 align:middle line:90% and unemployment. 00:07:00.160 --> 00:07:05.150 align:middle line:84% Inflation is the rate of change of P, no? 00:07:05.150 --> 00:07:06.740 align:middle line:90% It's not the level of P. 00:07:06.740 --> 00:07:09.590 align:middle line:84% So to do that, all that we'll do is-- so 00:07:09.590 --> 00:07:11.690 align:middle line:84% this, when I don't have a subscript here, 00:07:11.690 --> 00:07:16.520 align:middle line:84% I mean the price at time t, And this is the expected price 00:07:16.520 --> 00:07:17.270 align:middle line:90% for next period. 00:07:17.270 --> 00:07:18.840 align:middle line:90% That's what you have. 00:07:18.840 --> 00:07:21.140 align:middle line:90% But today, for the next period. 00:07:21.140 --> 00:07:23.450 align:middle line:84% What I'm going to do, I'm going to divide both sides 00:07:23.450 --> 00:07:24.770 align:middle line:90% by P minus 1. 00:07:24.770 --> 00:07:28.230 align:middle line:84% By that, I mean the price in the previous period. 00:07:28.230 --> 00:07:29.090 align:middle line:90% So both sides. 00:07:29.090 --> 00:07:33.350 align:middle line:84% I'm going to divide this side by P minus 1 and this one 00:07:33.350 --> 00:07:34.490 align:middle line:90% by P minus 1. 00:07:34.490 --> 00:07:36.800 align:middle line:90% So I get that expression. 00:07:36.800 --> 00:07:39.590 align:middle line:84% That's exactly the same equation we had before. 00:07:39.590 --> 00:07:42.680 align:middle line:84% All that I did is I divided by P minus 1. 00:07:42.680 --> 00:07:44.540 align:middle line:90% Remember what this means. 00:07:44.540 --> 00:07:49.840 align:middle line:84% So if this is the price for January 2023, 00:07:49.840 --> 00:07:52.210 align:middle line:84% this is the price for-- say, what, we're using 00:07:52.210 --> 00:07:55.390 align:middle line:90% annual data for January 2022? 00:07:55.390 --> 00:08:00.100 align:middle line:84% So I'm dividing by the price of January 2022 both sides. 00:08:00.100 --> 00:08:05.440 align:middle line:84% Now notice that-- remember that P over P minus 1 00:08:05.440 --> 00:08:07.960 align:middle line:84% is equal to 1 plus the inflation rate. 00:08:07.960 --> 00:08:10.540 align:middle line:84% Remember where inflation rate is just 00:08:10.540 --> 00:08:13.570 align:middle line:84% P minus P minus 1 over P minus 1. 00:08:13.570 --> 00:08:17.440 align:middle line:84% So this is just straightforward algebra, no? 00:08:17.440 --> 00:08:19.450 align:middle line:84% Remember our definition of inflation. 00:08:19.450 --> 00:08:33.880 align:middle line:90% 00:08:33.880 --> 00:08:36.159 align:middle line:84% That's P minus-- that's inflation. 00:08:36.159 --> 00:08:40.960 align:middle line:90% So 1 plus pi is just P minus-- 00:08:40.960 --> 00:08:42.700 align:middle line:90% over P minus 1. 00:08:42.700 --> 00:08:44.230 align:middle line:90% And that's what you have there. 00:08:44.230 --> 00:08:48.310 align:middle line:84% I can do the same for expected inflation. 00:08:48.310 --> 00:08:55.330 align:middle line:84% Notice that sometimes, people get confused. 00:08:55.330 --> 00:09:02.140 align:middle line:84% But expected inflation is equal to P expected, 00:09:02.140 --> 00:09:04.870 align:middle line:90% not minus P expected minus 1. 00:09:04.870 --> 00:09:07.800 align:middle line:90% It's P minus 1. 00:09:07.800 --> 00:09:10.140 align:middle line:90% P minus 1. 00:09:10.140 --> 00:09:12.810 align:middle line:84% And the reason I'm not subtracting the expectation 00:09:12.810 --> 00:09:15.870 align:middle line:84% here is because at time t, which is when you're 00:09:15.870 --> 00:09:17.880 align:middle line:84% forming that expectation, you already 00:09:17.880 --> 00:09:21.520 align:middle line:90% know what happened at t minus 1. 00:09:21.520 --> 00:09:23.970 align:middle line:84% So that's the reason this is expected inflation. 00:09:23.970 --> 00:09:30.760 align:middle line:84% I don't need to put expectations in here. 00:09:30.760 --> 00:09:32.860 align:middle line:90% So that's pi e. 00:09:32.860 --> 00:09:37.990 align:middle line:84% And so what we get is I can replace this guy 00:09:37.990 --> 00:09:42.520 align:middle line:84% here for 1 plus pi, this guy here for 1 plus pi e, 00:09:42.520 --> 00:09:44.920 align:middle line:84% and I get the following relationship. 00:09:44.920 --> 00:09:47.963 align:middle line:90% 00:09:47.963 --> 00:09:49.380 align:middle line:84% All that I've done is substituting 00:09:49.380 --> 00:09:52.100 align:middle line:90% this for that, that for that. 00:09:52.100 --> 00:09:55.070 align:middle line:84% So that's our price setting equation now expressed 00:09:55.070 --> 00:10:00.720 align:middle line:84% in terms of inflation rates and expected inflation rate. 00:10:00.720 --> 00:10:04.260 align:middle line:84% And now, we're not in Argentina, we're 00:10:04.260 --> 00:10:09.540 align:middle line:84% in the US, expected inflation are small numbers. 00:10:09.540 --> 00:10:11.760 align:middle line:84% And the log of 1 plus a small number 00:10:11.760 --> 00:10:13.050 align:middle line:90% is approximately that number. 00:10:13.050 --> 00:10:15.840 align:middle line:90% 00:10:15.840 --> 00:10:19.310 align:middle line:84% So I'm going to use this approximation, which, 00:10:19.310 --> 00:10:23.240 align:middle line:90% again, is valid for x small. 00:10:23.240 --> 00:10:28.415 align:middle line:84% And so I can replace this 1 plus pi for pi and this 1 plus pi 00:10:28.415 --> 00:10:34.430 align:middle line:84% e for pi e, this 1 plus m for m plus, 00:10:34.430 --> 00:10:38.060 align:middle line:84% and this term here, if these numbers are not too large, 00:10:38.060 --> 00:10:43.160 align:middle line:84% again, plus minus alpha u plus z. 00:10:43.160 --> 00:10:46.850 align:middle line:84% And I do all that and I end up with this expression. 00:10:46.850 --> 00:10:51.120 align:middle line:90% 00:10:51.120 --> 00:10:53.530 align:middle line:84% All I've done is I took logs of these. 00:10:53.530 --> 00:10:57.210 align:middle line:84% So I get log of 1 plus pi equal to log of 1 plus pi e 00:10:57.210 --> 00:11:02.640 align:middle line:84% plus log of 1 plus m plus log of 1 minus alpha u plus alpha z. 00:11:02.640 --> 00:11:07.410 align:middle line:84% I'm saying if pi, pi e, m, alpha u, plus z 00:11:07.410 --> 00:11:10.230 align:middle line:84% are not very large numbers, which we will assume, 00:11:10.230 --> 00:11:12.750 align:middle line:84% then this is approximately right. 00:11:12.750 --> 00:11:16.770 align:middle line:84% So I can rewrite that expression as that, approximately. 00:11:16.770 --> 00:11:18.520 align:middle line:84% I should have put in approximately. 00:11:18.520 --> 00:11:22.242 align:middle line:90% 00:11:22.242 --> 00:11:23.700 align:middle line:84% So now we have something that looks 00:11:23.700 --> 00:11:26.538 align:middle line:84% a lot more like the empirical relationship we 00:11:26.538 --> 00:11:27.330 align:middle line:90% were talking about. 00:11:27.330 --> 00:11:30.340 align:middle line:84% We have a relationship between inflation and unemployment. 00:11:30.340 --> 00:11:34.620 align:middle line:84% So this says that for any given expected inflation, and markups, 00:11:34.620 --> 00:11:38.710 align:middle line:84% and labor market institutions, higher unemployment 00:11:38.710 --> 00:11:40.240 align:middle line:90% means lower inflation. 00:11:40.240 --> 00:11:43.175 align:middle line:90% And why is that? 00:11:43.175 --> 00:11:45.550 align:middle line:84% So that curve tells you, that's the negative relationship 00:11:45.550 --> 00:11:46.290 align:middle line:90% we wanted, no? 00:11:46.290 --> 00:11:50.130 align:middle line:84% It says, higher unemployment, lower inflation. 00:11:50.130 --> 00:11:53.580 align:middle line:90% 00:11:53.580 --> 00:11:54.300 align:middle line:90% Why is that? 00:11:54.300 --> 00:12:02.070 align:middle line:90% 00:12:02.070 --> 00:12:05.430 align:middle line:84% Look, you heard it very clear when we talk about this, no? 00:12:05.430 --> 00:12:08.850 align:middle line:84% You understood very clearly why an increase in unemployment 00:12:08.850 --> 00:12:09.930 align:middle line:90% lowered the wage. 00:12:09.930 --> 00:12:14.580 align:middle line:84% You understood very clearly why, therefore, 00:12:14.580 --> 00:12:18.230 align:middle line:84% an increase in unemployment lowered the price. 00:12:18.230 --> 00:12:20.940 align:middle line:84% I haven't done anything but algebra in the two steps. 00:12:20.940 --> 00:12:24.380 align:middle line:84% So the same economics behind the explanations that you had before 00:12:24.380 --> 00:12:26.430 align:middle line:90% apply to this curve here. 00:12:26.430 --> 00:12:29.720 align:middle line:84% So the reason inflation will be lower 00:12:29.720 --> 00:12:33.200 align:middle line:84% when unemployment is higher, given all the rest, 00:12:33.200 --> 00:12:36.500 align:middle line:84% is because it will be less wage pressure. 00:12:36.500 --> 00:12:38.180 align:middle line:90% Workers will demand lower wages. 00:12:38.180 --> 00:12:40.400 align:middle line:84% That means lower prices, and therefore inflation 00:12:40.400 --> 00:12:41.120 align:middle line:90% will be lower. 00:12:41.120 --> 00:12:42.980 align:middle line:84% The economics doesn't change at all. 00:12:42.980 --> 00:12:47.600 align:middle line:84% I only divided both sides by P minus 1 and I took the logs 00:12:47.600 --> 00:12:48.810 align:middle line:90% and I approximated it. 00:12:48.810 --> 00:12:50.360 align:middle line:84% So the economics has not changed. 00:12:50.360 --> 00:12:55.620 align:middle line:84% I just did a little bit of basic math. 00:12:55.620 --> 00:12:58.433 align:middle line:84% So all-- what I'm trying to say is all the intuitions 00:12:58.433 --> 00:12:59.850 align:middle line:84% that you can already-- you already 00:12:59.850 --> 00:13:02.890 align:middle line:84% had from the wage setting, price setting equation 00:13:02.890 --> 00:13:05.850 align:middle line:84% and so on, you can apply to the Phillips curve as well. 00:13:05.850 --> 00:13:08.510 align:middle line:90% 00:13:08.510 --> 00:13:09.080 align:middle line:90% Good. 00:13:09.080 --> 00:13:13.400 align:middle line:84% So now we have something that, in principle, could explain 00:13:13.400 --> 00:13:16.940 align:middle line:84% the type of relationship that Phillips found and then 00:13:16.940 --> 00:13:20.720 align:middle line:84% Samuelson-Solow corroborated with extended data. 00:13:20.720 --> 00:13:26.570 align:middle line:90% 00:13:26.570 --> 00:13:27.070 align:middle line:90% Let's see. 00:13:27.070 --> 00:13:30.993 align:middle line:84% How do we get to something that looks like what these people run 00:13:30.993 --> 00:13:31.660 align:middle line:90% as a regression? 00:13:31.660 --> 00:13:34.420 align:middle line:84% Remember, they ran a regression essentially, 00:13:34.420 --> 00:13:40.690 align:middle line:84% or they correlated inflation with unemployment. 00:13:40.690 --> 00:13:44.100 align:middle line:84% And they found a downward sloping relationship. 00:13:44.100 --> 00:13:47.250 align:middle line:90% Well, look at what happens here. 00:13:47.250 --> 00:13:51.540 align:middle line:84% Suppose that we assume that the expected inflation is 00:13:51.540 --> 00:13:54.670 align:middle line:90% equal to some constant. 00:13:54.670 --> 00:13:57.850 align:middle line:84% In economics, we say when that's the case, 00:13:57.850 --> 00:14:00.250 align:middle line:84% and especially if pi is a low number, 00:14:00.250 --> 00:14:03.520 align:middle line:84% inflation expectations are well anchored, 00:14:03.520 --> 00:14:07.653 align:middle line:84% meaning any single year, there can be a price of oil is high, 00:14:07.653 --> 00:14:09.070 align:middle line:84% or something happens and inflation 00:14:09.070 --> 00:14:10.420 align:middle line:90% will deviate from that. 00:14:10.420 --> 00:14:13.390 align:middle line:84% But people are all the time expecting for inflation 00:14:13.390 --> 00:14:16.780 align:middle line:84% to go back to what is a normal level. 00:14:16.780 --> 00:14:22.150 align:middle line:84% Nowadays, or at least a few years ago, in the US, 00:14:22.150 --> 00:14:24.970 align:middle line:84% the normal level was around 2%, say. 00:14:24.970 --> 00:14:28.360 align:middle line:84% So people say, well, this year inflation was 1.8%, 00:14:28.360 --> 00:14:30.670 align:middle line:90% but we expect next year 2%. 00:14:30.670 --> 00:14:33.353 align:middle line:84% The next year, we got surprises on the upside. 00:14:33.353 --> 00:14:35.020 align:middle line:84% Price of food went up, or something like 00:14:35.020 --> 00:14:37.900 align:middle line:90% that, we got inflation of 2.3%. 00:14:37.900 --> 00:14:40.390 align:middle line:84% But you ask people how much you expect for next year 00:14:40.390 --> 00:14:42.730 align:middle line:90% and say, well, 2%. 00:14:42.730 --> 00:14:46.360 align:middle line:84% So that's what a model of expectation like this means. 00:14:46.360 --> 00:14:48.970 align:middle line:84% It's that you're always expecting something which 00:14:48.970 --> 00:14:51.670 align:middle line:84% is some historical value and that we 00:14:51.670 --> 00:14:54.560 align:middle line:84% have agreed is a reasonable level for our economy 00:14:54.560 --> 00:14:56.000 align:middle line:90% or something like that. 00:14:56.000 --> 00:14:59.420 align:middle line:84% So you see that if I replace expected inflation 00:14:59.420 --> 00:15:02.810 align:middle line:84% for a constant here by bar, then I have-- 00:15:02.810 --> 00:15:04.880 align:middle line:84% then my Phillips curve is really this. 00:15:04.880 --> 00:15:09.020 align:middle line:84% It's inflation, then I have a constant minus alpha u. 00:15:09.020 --> 00:15:12.020 align:middle line:84% That's the simplest of the downward sloping 00:15:12.020 --> 00:15:13.070 align:middle line:90% relationship I can have. 00:15:13.070 --> 00:15:15.800 align:middle line:84% That case is a downward sloping line. 00:15:15.800 --> 00:15:17.500 align:middle line:90% That's it. 00:15:17.500 --> 00:15:20.490 align:middle line:84% Of course, it could be nonlinear and so on. 00:15:20.490 --> 00:15:21.810 align:middle line:90% But this captures the essence. 00:15:21.810 --> 00:15:25.140 align:middle line:84% So that's a theory for why Phillips was 00:15:25.140 --> 00:15:27.930 align:middle line:90% finding what he was finding. 00:15:27.930 --> 00:15:31.620 align:middle line:84% Our theory of the wage of the labor market, if you will, 00:15:31.620 --> 00:15:34.830 align:middle line:84% and the price setting behavior of firms 00:15:34.830 --> 00:15:39.040 align:middle line:84% gives us a Phillips curve of the kind that he had in mind. 00:15:39.040 --> 00:15:43.080 align:middle line:84% And if you look in the '60s in the US, 00:15:43.080 --> 00:15:46.680 align:middle line:84% then you see this negative relationship that eventually 00:15:46.680 --> 00:15:47.600 align:middle line:90% became steeper. 00:15:47.600 --> 00:15:49.290 align:middle line:90% So it wasn't linear like this. 00:15:49.290 --> 00:15:53.590 align:middle line:84% It was a little convex, but it's downward sloping. 00:15:53.590 --> 00:16:02.410 align:middle line:84% And in fact, to some extent, our very own Bob Solow and Paul 00:16:02.410 --> 00:16:06.980 align:middle line:84% Samuelson were advising the US government at the time 00:16:06.980 --> 00:16:11.110 align:middle line:84% and they said, well, let's exploit this stuff a little. 00:16:11.110 --> 00:16:13.210 align:middle line:84% We like to have lower unemployment. 00:16:13.210 --> 00:16:15.430 align:middle line:84% We can live with a little more inflation, 00:16:15.430 --> 00:16:17.440 align:middle line:84% but we know that it's a negative trade off. 00:16:17.440 --> 00:16:19.910 align:middle line:84% It's a negative trade off between these two things. 00:16:19.910 --> 00:16:22.750 align:middle line:84% So if we'd like to lower unemployment, it's fine. 00:16:22.750 --> 00:16:24.250 align:middle line:84% We get a little more of inflation. 00:16:24.250 --> 00:16:26.320 align:middle line:84% And initially the deal was very good 00:16:26.320 --> 00:16:30.058 align:middle line:84% because this curve was very flat, you see? 00:16:30.058 --> 00:16:31.600 align:middle line:84% So you could cut unemployment a lot-- 00:16:31.600 --> 00:16:32.725 align:middle line:90% you can see the dates here. 00:16:32.725 --> 00:16:34.420 align:middle line:84% You're cutting unemployment a lot 00:16:34.420 --> 00:16:36.970 align:middle line:84% and you're not getting a lot of inflation. 00:16:36.970 --> 00:16:40.840 align:middle line:84% Eventually, the deal turned into a much rotten deal-- 00:16:40.840 --> 00:16:42.803 align:middle line:84% much more rotten deal, because then 00:16:42.803 --> 00:16:44.470 align:middle line:84% to lower a little bit more unemployment, 00:16:44.470 --> 00:16:46.840 align:middle line:84% we start getting a lot more inflation. 00:16:46.840 --> 00:16:50.907 align:middle line:84% So people, for a while, were OK with this model, 00:16:50.907 --> 00:16:52.240 align:middle line:90% assuming that inflation was low. 00:16:52.240 --> 00:16:55.960 align:middle line:84% But when they realized that this thing was being exploited, 00:16:55.960 --> 00:16:59.490 align:middle line:84% then they began to change the expectations they made. 00:16:59.490 --> 00:17:01.040 align:middle line:90% I think that's what we had here. 00:17:01.040 --> 00:17:04.533 align:middle line:84% But the region held pretty well during this time. 00:17:04.533 --> 00:17:06.200 align:middle line:84% And again, it became steeper and steeper 00:17:06.200 --> 00:17:10.220 align:middle line:84% as we pushed it more and more towards very low levels 00:17:10.220 --> 00:17:12.577 align:middle line:90% of unemployment. 00:17:12.577 --> 00:17:13.410 align:middle line:90% So that's the story. 00:17:13.410 --> 00:17:15.618 align:middle line:84% But again, there is your model of the Phillips curve. 00:17:15.618 --> 00:17:19.680 align:middle line:84% And that is a very good model for the times where Phillips 00:17:19.680 --> 00:17:23.750 align:middle line:90% also estimated his Phillips. 00:17:23.750 --> 00:17:28.730 align:middle line:84% Now if you turn the page and look at the same data 00:17:28.730 --> 00:17:33.520 align:middle line:90% in the '70s, look how it looks. 00:17:33.520 --> 00:17:36.970 align:middle line:84% So from 1970 to 1995, that's the data you have there, 00:17:36.970 --> 00:17:38.440 align:middle line:84% there is no negative relationship. 00:17:38.440 --> 00:17:41.120 align:middle line:90% I think it's all over the place. 00:17:41.120 --> 00:17:44.810 align:middle line:84% So had Mr Phillips been born a few years-- 00:17:44.810 --> 00:17:48.500 align:middle line:84% a few decades later and had he estimated his regression, 00:17:48.500 --> 00:17:50.780 align:middle line:90% he would have found nothing. 00:17:50.780 --> 00:17:52.610 align:middle line:84% There would be no curve in his honor, 00:17:52.610 --> 00:17:54.350 align:middle line:84% at least he had run that same regression. 00:17:54.350 --> 00:17:56.308 align:middle line:84% Maybe he would have run a different regression. 00:17:56.308 --> 00:17:58.010 align:middle line:90% But nothing. 00:17:58.010 --> 00:18:00.170 align:middle line:90% OK, so what happened? 00:18:00.170 --> 00:18:04.890 align:middle line:84% Well, our theory can explain as well what happened there. 00:18:04.890 --> 00:18:08.160 align:middle line:84% Remember the theory is not that inflation 00:18:08.160 --> 00:18:14.640 align:middle line:84% is equal to a constant minus alpha u. 00:18:14.640 --> 00:18:17.370 align:middle line:84% The theory says this is a constant 00:18:17.370 --> 00:18:21.610 align:middle line:84% only if the mode of expectation is this constant. 00:18:21.610 --> 00:18:24.270 align:middle line:84% But if the expectation is moving around, 00:18:24.270 --> 00:18:28.740 align:middle line:84% or if anything in this constant is moving around, 00:18:28.740 --> 00:18:31.830 align:middle line:84% then there is another source of variation. 00:18:31.830 --> 00:18:34.120 align:middle line:90% For example, what happens-- 00:18:34.120 --> 00:18:38.025 align:middle line:84% suppose you are here in 1965 and all of a sudden, you get-- 00:18:38.025 --> 00:18:39.910 align:middle line:90% the price of oil goes up a lot. 00:18:39.910 --> 00:18:41.970 align:middle line:84% And I'm telling you, capture the price of oil 00:18:41.970 --> 00:18:42.930 align:middle line:90% with an increase in m. 00:18:42.930 --> 00:18:45.210 align:middle line:84% Firms need to mark up things more in order 00:18:45.210 --> 00:18:47.760 align:middle line:90% to cover higher energy costs. 00:18:47.760 --> 00:18:49.800 align:middle line:90% Well, look at what m does. 00:18:49.800 --> 00:18:52.540 align:middle line:84% m says that for any given level of unemployment, 00:18:52.540 --> 00:18:54.040 align:middle line:90% now I get higher inflation. 00:18:54.040 --> 00:18:55.630 align:middle line:84% That's what an oil shock does, no? 00:18:55.630 --> 00:18:58.480 align:middle line:84% You get an oil shock, then for any given level of unemployment 00:18:58.480 --> 00:19:00.460 align:middle line:84% now, you find yourself with more inflation. 00:19:00.460 --> 00:19:02.820 align:middle line:84% So that moves you in the opposite direction. 00:19:02.820 --> 00:19:04.150 align:middle line:90% It moves you up there. 00:19:04.150 --> 00:19:08.600 align:middle line:84% And that's one of the reasons for these points around here. 00:19:08.600 --> 00:19:12.950 align:middle line:84% We got lots of inflation because we got massive oil shocks during 00:19:12.950 --> 00:19:14.770 align:middle line:90% the '70s and early '80s. 00:19:14.770 --> 00:19:18.720 align:middle line:90% 00:19:18.720 --> 00:19:21.180 align:middle line:84% We had wars in the Middle East and so on that 00:19:21.180 --> 00:19:23.280 align:middle line:90% led to those shocks. 00:19:23.280 --> 00:19:27.750 align:middle line:84% So that was one of the reasons we got shots here, 00:19:27.750 --> 00:19:29.340 align:middle line:90% to this term here. 00:19:29.340 --> 00:19:33.000 align:middle line:84% And that sort of muddied the relationship. 00:19:33.000 --> 00:19:36.300 align:middle line:84% But the other reason, which is more interesting, I think, 00:19:36.300 --> 00:19:39.180 align:middle line:84% and that you already began to see that something was happening 00:19:39.180 --> 00:19:44.020 align:middle line:84% here is that as inflation went up, 00:19:44.020 --> 00:19:46.400 align:middle line:84% people stopped believing in this model. 00:19:46.400 --> 00:19:49.885 align:middle line:84% So the expectation formation mechanism changed. 00:19:49.885 --> 00:19:52.810 align:middle line:90% 00:19:52.810 --> 00:19:58.520 align:middle line:84% So this guy began to react to endogenous variables. 00:19:58.520 --> 00:20:00.550 align:middle line:84% And I'm going to explain more precisely why. 00:20:00.550 --> 00:20:03.490 align:middle line:84% So that's what we mean by expecting 00:20:03.490 --> 00:20:05.350 align:middle line:90% inflation became deanchored. 00:20:05.350 --> 00:20:10.210 align:middle line:84% It was no longer anchored around this constant of 2%, 00:20:10.210 --> 00:20:11.530 align:middle line:90% but it became deanchored. 00:20:11.530 --> 00:20:13.190 align:middle line:90% It began to follow the data. 00:20:13.190 --> 00:20:15.010 align:middle line:84% So if the data came with more inflation, 00:20:15.010 --> 00:20:17.170 align:middle line:84% then people believed that next year, 00:20:17.170 --> 00:20:19.240 align:middle line:84% we would have more inflation as well. 00:20:19.240 --> 00:20:22.970 align:middle line:84% Not back to 2%, but if we got 5% inflation today, 00:20:22.970 --> 00:20:24.220 align:middle line:90% people began to say, well, OK. 00:20:24.220 --> 00:20:25.840 align:middle line:90% I don't think that next year-- 00:20:25.840 --> 00:20:27.580 align:middle line:90% my best estimate is 2%. 00:20:27.580 --> 00:20:29.878 align:middle line:90% It's probably closer to 5%. 00:20:29.878 --> 00:20:31.420 align:middle line:84% So that's what it means, deanchoring. 00:20:31.420 --> 00:20:34.840 align:middle line:84% That's what has the fed and most central banks around the world 00:20:34.840 --> 00:20:36.280 align:middle line:90% terrified today. 00:20:36.280 --> 00:20:40.000 align:middle line:84% Inflation is much higher than 2% and they're 00:20:40.000 --> 00:20:44.170 align:middle line:84% very worried about this guy becoming deanchored, 00:20:44.170 --> 00:20:46.330 align:middle line:90% or unanchored. 00:20:46.330 --> 00:20:48.280 align:middle line:84% I'll get back to that in a second. 00:20:48.280 --> 00:20:52.550 align:middle line:84% Anyway, but let me explain how this expected inflation 00:20:52.550 --> 00:20:54.220 align:middle line:90% term works. 00:20:54.220 --> 00:20:59.760 align:middle line:84% So let me replace the model of expected inflation for something 00:20:59.760 --> 00:21:05.640 align:middle line:84% which is some weighted average of a constant and the most 00:21:05.640 --> 00:21:08.050 align:middle line:90% recent inflation. 00:21:08.050 --> 00:21:10.840 align:middle line:84% So this model says, what is my expected inflation 00:21:10.840 --> 00:21:11.510 align:middle line:90% for next year? 00:21:11.510 --> 00:21:13.840 align:middle line:84% Well, it's an average of this long run target 00:21:13.840 --> 00:21:17.560 align:middle line:84% that we have, say, 2%, and whatever 00:21:17.560 --> 00:21:20.650 align:middle line:90% was the most recent inflation. 00:21:20.650 --> 00:21:23.320 align:middle line:84% If theta-- in the model I showed you before, 00:21:23.320 --> 00:21:25.840 align:middle line:84% the one that applied to the '60s and so on-- 00:21:25.840 --> 00:21:29.710 align:middle line:84% up to the '60s had essentially theta equal to 0. 00:21:29.710 --> 00:21:34.870 align:middle line:84% So this guy didn't show up there and expected inflation 00:21:34.870 --> 00:21:36.710 align:middle line:90% was very well anchored. 00:21:36.710 --> 00:21:41.630 align:middle line:84% What, what began to happen as we began to move that way 00:21:41.630 --> 00:21:43.670 align:middle line:84% and then we got hit by oil shocks, 00:21:43.670 --> 00:21:45.740 align:middle line:84% so people began to see much higher inflation 00:21:45.740 --> 00:21:49.880 align:middle line:84% numbers than they were used to, then this theta 00:21:49.880 --> 00:21:52.830 align:middle line:90% began to increase. 00:21:52.830 --> 00:21:55.700 align:middle line:84% So people began to change the model of expectation 00:21:55.700 --> 00:21:58.040 align:middle line:84% and began to think that the inflation was 00:21:58.040 --> 00:22:01.200 align:middle line:84% going to be more persistent than they used to think in the past. 00:22:01.200 --> 00:22:03.740 align:middle line:84% So high inflation today means high inflation tomorrow. 00:22:03.740 --> 00:22:05.480 align:middle line:84% That's what it means more persistent. 00:22:05.480 --> 00:22:07.550 align:middle line:84% In the past, it was high inflation today. 00:22:07.550 --> 00:22:08.720 align:middle line:90% That was a backdraw. 00:22:08.720 --> 00:22:11.750 align:middle line:84% We'll go back to the normal long run average. 00:22:11.750 --> 00:22:14.070 align:middle line:90% Now that's no longer the case. 00:22:14.070 --> 00:22:16.160 align:middle line:84% And so if I replace this more general model 00:22:16.160 --> 00:22:19.610 align:middle line:84% of expected inflation here in the Phillips curve, 00:22:19.610 --> 00:22:23.280 align:middle line:84% I get this expression, which now has this extra term. 00:22:23.280 --> 00:22:26.360 align:middle line:84% So we used to have theta equal to 0. 00:22:26.360 --> 00:22:30.470 align:middle line:84% But during the '70s and '80s, and even early '90s, actually 00:22:30.470 --> 00:22:34.280 align:middle line:84% that theta got to be very close to 1. 00:22:34.280 --> 00:22:36.050 align:middle line:84% You estimate these models, you get 00:22:36.050 --> 00:22:37.910 align:middle line:90% that theta was very close to 1. 00:22:37.910 --> 00:22:40.940 align:middle line:84% And look at what happens when theta gets very close to 1. 00:22:40.940 --> 00:22:44.720 align:middle line:84% So when theta is 1, literally, then the best forecast 00:22:44.720 --> 00:22:47.650 align:middle line:84% for inflation is the previous inflation. 00:22:47.650 --> 00:22:51.070 align:middle line:84% So this year is 5%, and I think next year is 5%. 00:22:51.070 --> 00:22:53.170 align:middle line:90% Not 2%, 5%. 00:22:53.170 --> 00:22:57.780 align:middle line:84% If this year is 7%, I think next year is 7% again. 00:22:57.780 --> 00:23:00.860 align:middle line:84% And so if you do that, then my expected inflation 00:23:00.860 --> 00:23:03.080 align:middle line:84% becomes lag inflation by t minus 1. 00:23:03.080 --> 00:23:07.460 align:middle line:84% So if I replace the expected inflation for pi t minus 1, 00:23:07.460 --> 00:23:09.710 align:middle line:84% I get to this Phillips curve, which 00:23:09.710 --> 00:23:14.540 align:middle line:84% I can rewrite as the change in inflation-- 00:23:14.540 --> 00:23:17.990 align:middle line:84% as a relationship between the change in inflation 00:23:17.990 --> 00:23:20.130 align:middle line:90% and the level of unemployment. 00:23:20.130 --> 00:23:24.060 align:middle line:84% So now what you have is that if unemployment is very low, 00:23:24.060 --> 00:23:27.500 align:middle line:90% then inflation is picking up. 00:23:27.500 --> 00:23:32.030 align:middle line:84% It's going-- so if inflation-- if unemployment is very low, not 00:23:32.030 --> 00:23:37.100 align:middle line:84% only the inflation is high, but it's also growing over time. 00:23:37.100 --> 00:23:39.170 align:middle line:84% That's the reason sometimes people 00:23:39.170 --> 00:23:41.810 align:middle line:84% refer to this formulation of the Phillips curve 00:23:41.810 --> 00:23:44.617 align:middle line:84% as the accelerationist Phillips curve, 00:23:44.617 --> 00:23:46.700 align:middle line:84% because now it's the relation between unemployment 00:23:46.700 --> 00:23:48.720 align:middle line:90% and the change in inflation. 00:23:48.720 --> 00:23:51.800 align:middle line:84% And if you estimate this Phillips curve, 00:23:51.800 --> 00:23:55.220 align:middle line:84% this accelerationist Phillips curve on the data I just showed 00:23:55.220 --> 00:24:00.260 align:middle line:84% you of the '70s and '80s, you get a much better relationship. 00:24:00.260 --> 00:24:03.560 align:middle line:84% You still have the oil shocks that mess things up, 00:24:03.560 --> 00:24:06.390 align:middle line:84% but you start recovering this negative relationship. 00:24:06.390 --> 00:24:08.910 align:middle line:84% But again, it's between the change in inflation 00:24:08.910 --> 00:24:10.160 align:middle line:90% and the level of unemployment. 00:24:10.160 --> 00:24:12.560 align:middle line:84% And that's a very scary situation for the central bank 00:24:12.560 --> 00:24:15.260 align:middle line:84% to find itself in, because it's very easy for things 00:24:15.260 --> 00:24:16.100 align:middle line:90% to escalate. 00:24:16.100 --> 00:24:22.270 align:middle line:90% 00:24:22.270 --> 00:24:27.460 align:middle line:84% So by the mid-90s, we had reanchored expectations. 00:24:27.460 --> 00:24:30.580 align:middle line:84% There were very aggressive policies 00:24:30.580 --> 00:24:35.620 align:middle line:84% to control inflation by Paul Volcker in the US 00:24:35.620 --> 00:24:39.670 align:middle line:84% and it was imitated around the world with some lag. 00:24:39.670 --> 00:24:41.650 align:middle line:90% But inflation became reanchored. 00:24:41.650 --> 00:24:45.430 align:middle line:84% So we went back to this theta equal to 0 type model. 00:24:45.430 --> 00:24:47.020 align:middle line:84% The expected inflation in the US-- 00:24:47.020 --> 00:24:50.380 align:middle line:84% the target inflation of the central bank was around 2%. 00:24:50.380 --> 00:24:54.310 align:middle line:84% That became what people expected for the next year 00:24:54.310 --> 00:25:00.280 align:middle line:84% and that reanchored-- so we went back, in other words, 00:25:00.280 --> 00:25:03.520 align:middle line:90% to that Phillips curve. 00:25:03.520 --> 00:25:05.660 align:middle line:84% And that's what central banks want to be at. 00:25:05.660 --> 00:25:08.830 align:middle line:84% They want to have inflation expectations very well anchored. 00:25:08.830 --> 00:25:11.350 align:middle line:84% And they were very successful after the '90s. 00:25:11.350 --> 00:25:14.620 align:middle line:84% And so we got into again-- now look. 00:25:14.620 --> 00:25:16.870 align:middle line:84% And now I'm not running the accelerationist-- 00:25:16.870 --> 00:25:19.360 align:middle line:84% I'm again running inflation against unemployment 00:25:19.360 --> 00:25:24.070 align:middle line:84% and you again see this downward sloping relationship. 00:25:24.070 --> 00:25:25.540 align:middle line:90% So that was very good news. 00:25:25.540 --> 00:25:28.990 align:middle line:84% It was a great success of monetary policy during 00:25:28.990 --> 00:25:32.920 align:middle line:84% the '90s and later was the reanchoring of expected 00:25:32.920 --> 00:25:35.870 align:middle line:84% inflation again all around the developed world, 00:25:35.870 --> 00:25:36.910 align:middle line:90% and many of the-- 00:25:36.910 --> 00:25:38.800 align:middle line:90% even Latin America. 00:25:38.800 --> 00:25:41.620 align:middle line:84% Many economists in Latin America reanchored their expectations, 00:25:41.620 --> 00:25:43.150 align:middle line:90% Asia, and so on. 00:25:43.150 --> 00:25:45.100 align:middle line:84% So it was a good time for central banks. 00:25:45.100 --> 00:25:58.840 align:middle line:90% 00:25:58.840 --> 00:26:02.170 align:middle line:84% So the next thing I want to do, this will connect more 00:26:02.170 --> 00:26:04.420 align:middle line:84% with the previous lecture, and it's 00:26:04.420 --> 00:26:09.390 align:middle line:84% the last thing I want to say for this lecture, 00:26:09.390 --> 00:26:13.730 align:middle line:84% and I may start the review afterwards, 00:26:13.730 --> 00:26:16.610 align:middle line:84% is that I want to connect now this Phillips 00:26:16.610 --> 00:26:17.995 align:middle line:84% curve with something we discussed 00:26:17.995 --> 00:26:19.370 align:middle line:84% in the previous lecture, which is 00:26:19.370 --> 00:26:20.662 align:middle line:90% a natural rate of unemployment. 00:26:20.662 --> 00:26:23.510 align:middle line:84% Because that's the way you typically see the Phillips 00:26:23.510 --> 00:26:27.380 align:middle line:84% curve written and that's also the way 00:26:27.380 --> 00:26:32.570 align:middle line:84% that when Chairman Powell is talking about the labor market 00:26:32.570 --> 00:26:34.910 align:middle line:84% tightness and so on, he's not talking 00:26:34.910 --> 00:26:37.017 align:middle line:84% relative to m, and z, and things like that. 00:26:37.017 --> 00:26:39.350 align:middle line:84% He's talking relative to what is called the natural rate 00:26:39.350 --> 00:26:40.110 align:middle line:90% of unemployment. 00:26:40.110 --> 00:26:42.350 align:middle line:84% So I want to go from a Phillips curve that 00:26:42.350 --> 00:26:49.860 align:middle line:84% looks like that the one that has the natural rate of unemployment 00:26:49.860 --> 00:26:50.740 align:middle line:90% in there. 00:26:50.740 --> 00:26:54.630 align:middle line:84% And so that's the last step in this lecture. 00:26:54.630 --> 00:26:57.440 align:middle line:84% So remember the definition of the natural rate 00:26:57.440 --> 00:26:58.460 align:middle line:90% of unemployment? 00:26:58.460 --> 00:27:01.040 align:middle line:84% What was the definition of the natural rate of unemployment? 00:27:01.040 --> 00:27:04.360 align:middle line:90% 00:27:04.360 --> 00:27:11.250 align:middle line:84% Was it the unemployment rate that God gave us, any God? 00:27:11.250 --> 00:27:15.380 align:middle line:90% 00:27:15.380 --> 00:27:17.540 align:middle line:90% No. 00:27:17.540 --> 00:27:19.280 align:middle line:84% It had a very precise meaning for us. 00:27:19.280 --> 00:27:23.880 align:middle line:90% 00:27:23.880 --> 00:27:30.250 align:middle line:84% And remember, we used exactly that model to figure it out. 00:27:30.250 --> 00:27:30.750 align:middle line:90% Remember? 00:27:30.750 --> 00:27:34.240 align:middle line:90% 00:27:34.240 --> 00:27:37.990 align:middle line:84% We saw-- actually, we saw the natural rate of unemployment 00:27:37.990 --> 00:27:39.160 align:middle line:90% from something like this. 00:27:39.160 --> 00:27:42.520 align:middle line:84% I think we had-- the function is still generic function F of u z, 00:27:42.520 --> 00:27:46.170 align:middle line:84% but we solved from an expression like this. 00:27:46.170 --> 00:27:50.070 align:middle line:84% And we said, under one assumption, 00:27:50.070 --> 00:27:55.190 align:middle line:84% we can call this u, un, the natural rate of unemployment. 00:27:55.190 --> 00:27:56.890 align:middle line:90% What was that assumption? 00:27:56.890 --> 00:27:58.015 align:middle line:90% And that's the only thing-- 00:27:58.015 --> 00:28:00.425 align:middle line:84% STUDENT: The price is equal to the expected price. 00:28:00.425 --> 00:28:01.800 align:middle line:84% RICARDO CABALLERO: Expected price 00:28:01.800 --> 00:28:03.520 align:middle line:90% is equal to the actual price. 00:28:03.520 --> 00:28:05.460 align:middle line:84% So we said, if this is equal to that, 00:28:05.460 --> 00:28:08.130 align:middle line:84% then you solve out the natural rate of unemployment. 00:28:08.130 --> 00:28:13.530 align:middle line:84% And that's the only thing that natural rate of unemployment 00:28:13.530 --> 00:28:16.650 align:middle line:84% means, simply that when the price is 00:28:16.650 --> 00:28:20.110 align:middle line:90% equal to the expected price. 00:28:20.110 --> 00:28:27.190 align:middle line:84% But if the price is equal to expected price, what 00:28:27.190 --> 00:28:28.023 align:middle line:90% else is equal? 00:28:28.023 --> 00:28:32.753 align:middle line:90% 00:28:32.753 --> 00:28:35.936 align:middle line:84% And I pointed at the right expressions. 00:28:35.936 --> 00:28:37.650 align:middle line:90% STUDENT: [INAUDIBLE]. 00:28:37.650 --> 00:28:40.950 align:middle line:84% RICARDO CABALLERO: Inflation is equal to expected inflation. 00:28:40.950 --> 00:28:44.350 align:middle line:84% So I can use the same logic I used here 00:28:44.350 --> 00:28:45.850 align:middle line:84% for the natural rate of unemployment 00:28:45.850 --> 00:28:47.740 align:middle line:90% using the Phillips curve. 00:28:47.740 --> 00:28:51.490 align:middle line:84% I can say, OK, I can solve out for the natural rate 00:28:51.490 --> 00:28:55.030 align:middle line:84% of unemployment here simply by setting the expected inflation 00:28:55.030 --> 00:28:56.470 align:middle line:90% equal to actual inflation. 00:28:56.470 --> 00:28:59.730 align:middle line:90% 00:28:59.730 --> 00:29:02.250 align:middle line:84% And if I do this, I can solve for the natural rate 00:29:02.250 --> 00:29:05.100 align:middle line:90% of unemployment from here, un. 00:29:05.100 --> 00:29:06.330 align:middle line:90% I mean, I'm going to give-- 00:29:06.330 --> 00:29:08.550 align:middle line:84% I'm going to put the superscript n here when 00:29:08.550 --> 00:29:10.695 align:middle line:90% you let me replace pi for pi. 00:29:10.695 --> 00:29:13.830 align:middle line:90% 00:29:13.830 --> 00:29:16.170 align:middle line:84% The fact that I replaced this pi e for pi 00:29:16.170 --> 00:29:18.482 align:middle line:84% is what allows me to put the superscript n there. 00:29:18.482 --> 00:29:20.190 align:middle line:84% Call it the natural rate of unemployment. 00:29:20.190 --> 00:29:21.287 align:middle line:90% And now I can solve it. 00:29:21.287 --> 00:29:22.870 align:middle line:84% Well, obviously that cancels with that 00:29:22.870 --> 00:29:24.960 align:middle line:84% and I can solve the natural rate of unemployment, 00:29:24.960 --> 00:29:28.570 align:middle line:84% and it's equal to this function here. 00:29:28.570 --> 00:29:33.470 align:middle line:84% So why is the natural rate of unemployment increasing m? 00:29:33.470 --> 00:29:37.905 align:middle line:84% A question like that can come up in the quiz. 00:29:37.905 --> 00:29:40.030 align:middle line:84% I'm not going to use the Phillips curve to ask you, 00:29:40.030 --> 00:29:41.030 align:middle line:90% if I ask you about that. 00:29:41.030 --> 00:29:42.500 align:middle line:90% But I can ask you that. 00:29:42.500 --> 00:29:47.350 align:middle line:84% What happens to a natural rate of unemployment if m goes up? 00:29:47.350 --> 00:29:51.550 align:middle line:84% You know that un will go up, but what is the mechanism? 00:29:51.550 --> 00:29:59.190 align:middle line:90% 00:29:59.190 --> 00:30:01.040 align:middle line:84% So why does the natural rate of unemployment 00:30:01.040 --> 00:30:05.830 align:middle line:90% go up when the markup goes up? 00:30:05.830 --> 00:30:06.330 align:middle line:90% Yep. 00:30:06.330 --> 00:30:08.288 align:middle line:84% STUDENT: [INAUDIBLE] our main cause and we just 00:30:08.288 --> 00:30:09.727 align:middle line:90% have to go down, right? 00:30:09.727 --> 00:30:11.310 align:middle line:84% RICARDO CABALLERO: I mean, another way 00:30:11.310 --> 00:30:14.040 align:middle line:84% of saying it is that the firms are not willing to pay-- they 00:30:14.040 --> 00:30:16.820 align:middle line:90% want to pay a lower real wage. 00:30:16.820 --> 00:30:19.580 align:middle line:84% At the original level of unemployment 00:30:19.580 --> 00:30:23.030 align:middle line:84% before the change in m, workers would not 00:30:23.030 --> 00:30:26.418 align:middle line:90% take that lower real wage. 00:30:26.418 --> 00:30:28.710 align:middle line:84% It's not an equilibrium real wage, because workers say, 00:30:28.710 --> 00:30:29.210 align:middle line:90% no, no. 00:30:29.210 --> 00:30:32.790 align:middle line:84% At this level of unemployment, we need a higher real wage. 00:30:32.790 --> 00:30:36.080 align:middle line:84% So the only way to restore equilibrium in that model we had 00:30:36.080 --> 00:30:38.480 align:middle line:84% was to increase unemployment, because that 00:30:38.480 --> 00:30:40.610 align:middle line:84% will lower the bargaining power of workers, 00:30:40.610 --> 00:30:43.670 align:middle line:84% and they will end up accepting the lower real wage that firms 00:30:43.670 --> 00:30:46.040 align:middle line:90% are willing to offer now. 00:30:46.040 --> 00:30:50.200 align:middle line:84% So that's the reason and we get this markup effect. 00:30:50.200 --> 00:30:53.280 align:middle line:90% 00:30:53.280 --> 00:30:56.328 align:middle line:90% z is same logic. 00:30:56.328 --> 00:30:57.870 align:middle line:84% It's a little easier to see it there, 00:30:57.870 --> 00:31:02.250 align:middle line:84% but z means, well, at any given level of unemployment, 00:31:02.250 --> 00:31:06.060 align:middle line:84% an increase in z means workers want a higher real wage. 00:31:06.060 --> 00:31:09.010 align:middle line:84% Firms are not willing to pay a higher real wage. 00:31:09.010 --> 00:31:12.510 align:middle line:84% So you have to bring down the real wage and workers' demand, 00:31:12.510 --> 00:31:15.887 align:middle line:84% and the only way that can happen is with a higher unemployment. 00:31:15.887 --> 00:31:17.970 align:middle line:84% That's the reason the natural rate of unemployment 00:31:17.970 --> 00:31:20.100 align:middle line:90% is also increasing in z. 00:31:20.100 --> 00:31:24.130 align:middle line:90% 00:31:24.130 --> 00:31:26.830 align:middle line:90% And now the last step. 00:31:26.830 --> 00:31:29.380 align:middle line:90% The last step is to-- 00:31:29.380 --> 00:31:36.890 align:middle line:84% you see, I can go back to my Philips curve, say that. 00:31:36.890 --> 00:31:42.950 align:middle line:84% And I'm going to replace m plus z for alpha un. 00:31:42.950 --> 00:31:45.110 align:middle line:90% I can do that, you see? 00:31:45.110 --> 00:31:49.670 align:middle line:84% I can replace this m plus z for alpha times un. 00:31:49.670 --> 00:31:52.598 align:middle line:90% 00:31:52.598 --> 00:31:53.390 align:middle line:90% How do I know that? 00:31:53.390 --> 00:31:58.720 align:middle line:84% Well, m plus z, z is equal to un times alpha. 00:31:58.720 --> 00:32:02.890 align:middle line:84% We can replace in the Phillips curve m plus c by alpha un 00:32:02.890 --> 00:32:07.690 align:middle line:84% and I can therefore rewrite the Phillips curve 00:32:07.690 --> 00:32:10.120 align:middle line:90% in the following form. 00:32:10.120 --> 00:32:12.310 align:middle line:84% Inflation is equal to expected inflation 00:32:12.310 --> 00:32:17.470 align:middle line:84% minus alpha times the gap between the unemployment rate 00:32:17.470 --> 00:32:19.330 align:middle line:84% and the natural rate of unemployment. 00:32:19.330 --> 00:32:22.110 align:middle line:90% 00:32:22.110 --> 00:32:28.560 align:middle line:84% So when Chairman Powell is worried about labor market very 00:32:28.560 --> 00:32:31.410 align:middle line:84% tight, what he's saying is, well, unemployment 00:32:31.410 --> 00:32:34.970 align:middle line:84% is likely to be below the natural rate of unemployment. 00:32:34.970 --> 00:32:37.070 align:middle line:84% Because if unemployment is below the natural rate 00:32:37.070 --> 00:32:39.740 align:middle line:84% of unemployment, that's putting upward pressure on inflation. 00:32:39.740 --> 00:32:44.070 align:middle line:90% 00:32:44.070 --> 00:32:46.230 align:middle line:84% So that's a-- so that's what he means. 00:32:46.230 --> 00:32:51.090 align:middle line:84% This gap is very important for macroeconomists, and certainly 00:32:51.090 --> 00:32:54.510 align:middle line:84% for central bankers that are very worried about inflation, 00:32:54.510 --> 00:32:55.500 align:middle line:90% that gap here. 00:32:55.500 --> 00:32:59.170 align:middle line:84% Problem is, is this is a difficult object to estimate. 00:32:59.170 --> 00:33:01.890 align:middle line:90% So you have to have estimates. 00:33:01.890 --> 00:33:05.035 align:middle line:84% The truth is that it's very difficult to know what it is, 00:33:05.035 --> 00:33:06.660 align:middle line:84% although there are estimates out there, 00:33:06.660 --> 00:33:08.930 align:middle line:90% and I'm going to show you one. 00:33:08.930 --> 00:33:10.670 align:middle line:84% You notice that something is wrong when 00:33:10.670 --> 00:33:12.680 align:middle line:90% this guy starts speaking up. 00:33:12.680 --> 00:33:16.545 align:middle line:84% It's a little bit the other way around. 00:33:16.545 --> 00:33:23.000 align:middle line:84% The US, in fact, had the opposite problem before COVID. 00:33:23.000 --> 00:33:24.530 align:middle line:84% It's that somehow, unemployment was 00:33:24.530 --> 00:33:26.300 align:middle line:84% very low, relative to historical levels, 00:33:26.300 --> 00:33:28.370 align:middle line:84% but inflation was not picking up. 00:33:28.370 --> 00:33:30.500 align:middle line:84% So that was implicitly telling us 00:33:30.500 --> 00:33:33.200 align:middle line:84% that for some reason not fully understood, 00:33:33.200 --> 00:33:37.990 align:middle line:84% the natural rate of unemployment was declining. 00:33:37.990 --> 00:33:40.720 align:middle line:84% So here is one picture that looks-- 00:33:40.720 --> 00:33:42.760 align:middle line:90% is one estimate. 00:33:42.760 --> 00:33:44.890 align:middle line:84% Again, I don't trust any particular estimate, 00:33:44.890 --> 00:33:47.170 align:middle line:90% but it tells a story. 00:33:47.170 --> 00:33:49.510 align:middle line:84% That's one particular estimate of the natural rate 00:33:49.510 --> 00:33:52.190 align:middle line:84% of unemployment in the US, that blue line. 00:33:52.190 --> 00:33:55.120 align:middle line:84% And what you see in red, the red is the actual rate 00:33:55.120 --> 00:33:57.390 align:middle line:90% of unemployment in the US. 00:33:57.390 --> 00:34:01.110 align:middle line:84% So what happens in situations like this? 00:34:01.110 --> 00:34:04.660 align:middle line:90% 00:34:04.660 --> 00:34:06.810 align:middle line:84% So what do you think was happening to inflation 00:34:06.810 --> 00:34:09.750 align:middle line:84% in this episode, which is right after the global financial 00:34:09.750 --> 00:34:11.880 align:middle line:90% crisis, or the Great Recession? 00:34:11.880 --> 00:34:15.380 align:middle line:90% 00:34:15.380 --> 00:34:17.400 align:middle line:84% So what do you need to read here? 00:34:17.400 --> 00:34:20.010 align:middle line:84% Well, the unemployment rate was a lot higher 00:34:20.010 --> 00:34:24.060 align:middle line:84% than the natural rate of unemployment. 00:34:24.060 --> 00:34:28.190 align:middle line:84% Does that put upward or downward pressure on inflation? 00:34:28.190 --> 00:34:29.510 align:middle line:90% Downward pressure on inflation. 00:34:29.510 --> 00:34:31.280 align:middle line:84% Unemployment is very high relative to natural rate 00:34:31.280 --> 00:34:32.179 align:middle line:90% of unemployment. 00:34:32.179 --> 00:34:35.288 align:middle line:84% It's minus alpha times u minus un. 00:34:35.288 --> 00:34:36.330 align:middle line:90% And that's what happened. 00:34:36.330 --> 00:34:37.955 align:middle line:84% We had lots of problems with inflation. 00:34:37.955 --> 00:34:39.290 align:middle line:90% Inflation was going very low. 00:34:39.290 --> 00:34:41.300 align:middle line:84% We even had negative inflation there, 00:34:41.300 --> 00:34:43.580 align:middle line:90% a little deflation for a while. 00:34:43.580 --> 00:34:46.719 align:middle line:90% So that was the problem. 00:34:46.719 --> 00:34:48.587 align:middle line:84% Here is the period that I described before. 00:34:48.587 --> 00:34:50.920 align:middle line:84% It's a little mysterious, because we went-- unemployment 00:34:50.920 --> 00:34:53.139 align:middle line:84% went below what we thought was a natural rate of unemployment 00:34:53.139 --> 00:34:55.090 align:middle line:84% and inflation wasn't really picking up a lot. 00:34:55.090 --> 00:34:56.798 align:middle line:84% At the end, it began to pick up a little. 00:34:56.798 --> 00:34:58.480 align:middle line:84% But it wasn't picking up a lot, and that 00:34:58.480 --> 00:35:00.740 align:middle line:90% was a little bit of a mystery. 00:35:00.740 --> 00:35:03.180 align:middle line:84% Now we're in this situation here, 00:35:03.180 --> 00:35:07.830 align:middle line:84% which we have extremely low unemployment 00:35:07.830 --> 00:35:09.980 align:middle line:90% and very high inflation. 00:35:09.980 --> 00:35:12.570 align:middle line:84% So I think this captures well the situation right now. 00:35:12.570 --> 00:35:15.510 align:middle line:84% We have a negative gap between unemployment 00:35:15.510 --> 00:35:17.310 align:middle line:84% and the natural rate of unemployment. 00:35:17.310 --> 00:35:19.560 align:middle line:84% And that's the reason that's putting a lot of pressure 00:35:19.560 --> 00:35:20.625 align:middle line:90% on inflation. 00:35:20.625 --> 00:35:22.750 align:middle line:84% We also have other things that are putting pressure 00:35:22.750 --> 00:35:25.167 align:middle line:84% on inflation that come from the supply side of the economy 00:35:25.167 --> 00:35:25.850 align:middle line:90% and so on. 00:35:25.850 --> 00:35:32.740 align:middle line:84% So that combination is pretty bad for the inflation outcomes 00:35:32.740 --> 00:35:35.710 align:middle line:90% and outlook as well. 00:35:35.710 --> 00:35:38.620 align:middle line:90% 00:35:38.620 --> 00:35:40.578 align:middle line:90% So that's where we're at. 00:35:40.578 --> 00:35:42.870 align:middle line:84% We're going to talk a lot more about this, because this 00:35:42.870 --> 00:35:45.400 align:middle line:90% is what is going on right now. 00:35:45.400 --> 00:35:46.443 align:middle line:90% Any questions about that? 00:35:46.443 --> 00:35:48.610 align:middle line:84% Otherwise, I want to start sort of reviewing things. 00:35:48.610 --> 00:35:51.630 align:middle line:90% Although, I don't know-- 00:35:51.630 --> 00:35:53.040 align:middle line:90% any question about this? 00:35:53.040 --> 00:35:53.540 align:middle line:90% Yeah? 00:35:53.540 --> 00:35:55.582 align:middle line:84% STUDENT: Is the only way to fix it this direction 00:35:55.582 --> 00:35:56.948 align:middle line:90% to increase unemployment? 00:35:56.948 --> 00:35:57.990 align:middle line:90% RICARDO CABALLERO: Sorry? 00:35:57.990 --> 00:36:01.120 align:middle line:84% STUDENT: Is the only way to fix, I guess, the inflationary-- 00:36:01.120 --> 00:36:03.933 align:middle line:84% RICARDO CABALLERO: Well, that's a very good question. 00:36:03.933 --> 00:36:05.100 align:middle line:90% That's a very good question. 00:36:05.100 --> 00:36:09.220 align:middle line:90% 00:36:09.220 --> 00:36:15.800 align:middle line:84% I'm trying to decide what to answer it with what we have. 00:36:15.800 --> 00:36:23.070 align:middle line:90% 00:36:23.070 --> 00:36:27.030 align:middle line:84% There are two views at this moment. 00:36:27.030 --> 00:36:32.560 align:middle line:84% There's one view that says there's no way around that. 00:36:32.560 --> 00:36:33.560 align:middle line:90% Just look at this curve. 00:36:33.560 --> 00:36:35.630 align:middle line:84% It says, look, there is no way around that. 00:36:35.630 --> 00:36:38.780 align:middle line:84% That's the reason we need a recession. 00:36:38.780 --> 00:36:43.360 align:middle line:84% Because otherwise, we're not going to control inflation. 00:36:43.360 --> 00:36:45.890 align:middle line:84% A recession means high unemployment. 00:36:45.890 --> 00:36:47.810 align:middle line:90% That's one view. 00:36:47.810 --> 00:36:51.440 align:middle line:84% At this moment, it's becoming the dominant view. 00:36:51.440 --> 00:36:52.670 align:middle line:90% It has gone in cycles. 00:36:52.670 --> 00:36:56.970 align:middle line:84% But at this moment, it's the dominant view. 00:36:56.970 --> 00:37:00.510 align:middle line:84% There is another view, which is the one 00:37:00.510 --> 00:37:04.560 align:middle line:84% that the central bank, the fed adopted for a while that 00:37:04.560 --> 00:37:07.710 align:middle line:84% said, well, this is not the only indicator of tightness 00:37:07.710 --> 00:37:08.670 align:middle line:90% of the labor market. 00:37:08.670 --> 00:37:11.120 align:middle line:90% There is other things as well. 00:37:11.120 --> 00:37:14.820 align:middle line:84% And those indicators are moving in the right direction. 00:37:14.820 --> 00:37:17.960 align:middle line:84% And so we may be able not to create a big mess here 00:37:17.960 --> 00:37:22.520 align:middle line:84% because these other factors are moving in the right direction. 00:37:22.520 --> 00:37:24.270 align:middle line:84% Some of those factors are, as I said, 00:37:24.270 --> 00:37:28.343 align:middle line:84% other measures of labor market tightness and hiring, the flows. 00:37:28.343 --> 00:37:30.260 align:middle line:84% Remember I showed you flows between employment 00:37:30.260 --> 00:37:32.218 align:middle line:84% and unemployment, out of employment, and so on? 00:37:32.218 --> 00:37:35.780 align:middle line:84% Those flows look extremely tight and now they're improving. 00:37:35.780 --> 00:37:37.970 align:middle line:84% So the gaps in those dimensions are better. 00:37:37.970 --> 00:37:41.150 align:middle line:84% And the other one is there was a big cost push component, which 00:37:41.150 --> 00:37:42.080 align:middle line:90% is what I said before. 00:37:42.080 --> 00:37:44.150 align:middle line:84% The supply chains and so on created 00:37:44.150 --> 00:37:46.160 align:middle line:84% extra inflation, abnormal inflation, 00:37:46.160 --> 00:37:47.840 align:middle line:90% like increasing markups. 00:37:47.840 --> 00:37:49.040 align:middle line:90% m was very high. 00:37:49.040 --> 00:37:51.320 align:middle line:84% And some of that is subsiding as well. 00:37:51.320 --> 00:37:53.120 align:middle line:84% So there are dynamics that suggest 00:37:53.120 --> 00:37:55.980 align:middle line:84% that inflation is declining even without unemployment. 00:37:55.980 --> 00:38:02.480 align:middle line:84% But I would say the median voter in this space of forecasts, 00:38:02.480 --> 00:38:05.570 align:middle line:84% of inflation and so on thinks that we 00:38:05.570 --> 00:38:11.050 align:middle line:84% will need some adjustment through this part as well. 00:38:11.050 --> 00:38:13.570 align:middle line:90% My main concern-- 00:38:13.570 --> 00:38:19.570 align:middle line:84% I think that the path the fed is forecasting is feasible, 00:38:19.570 --> 00:38:21.010 align:middle line:90% but a very narrow path. 00:38:21.010 --> 00:38:22.510 align:middle line:90% I mean, it may happen. 00:38:22.510 --> 00:38:26.320 align:middle line:84% And to me, whether they're successful at not creating 00:38:26.320 --> 00:38:29.590 align:middle line:84% a big mess here, I mean bringing unemployment very high 00:38:29.590 --> 00:38:31.990 align:middle line:84% in order to bring inflation down, 00:38:31.990 --> 00:38:34.780 align:middle line:84% has a lot to do with whether somehow we 00:38:34.780 --> 00:38:37.860 align:middle line:84% manage to keep expected inflation anchored. 00:38:37.860 --> 00:38:39.930 align:middle line:90% And there was some evidence-- 00:38:39.930 --> 00:38:41.670 align:middle line:84% I think I said that a few lectures ago. 00:38:41.670 --> 00:38:48.395 align:middle line:84% There was some evidence that in the summer of 2022. 00:38:48.395 --> 00:38:49.770 align:middle line:84% I'm from the Southern hemisphere, 00:38:49.770 --> 00:38:52.410 align:middle line:84% so I get always confused with summers and so on. 00:38:52.410 --> 00:38:56.930 align:middle line:84% So in the summer of 2022, US summer of 2022, inflation 00:38:56.930 --> 00:38:58.160 align:middle line:90% was becoming very unanchored. 00:38:58.160 --> 00:39:02.480 align:middle line:84% This guy, one year expected inflation was creeping up to 6%, 00:39:02.480 --> 00:39:04.430 align:middle line:90% and that was very scary. 00:39:04.430 --> 00:39:06.420 align:middle line:90% Because think what happened. 00:39:06.420 --> 00:39:10.160 align:middle line:84% If you get expected inflation at 6%, 00:39:10.160 --> 00:39:13.550 align:middle line:84% then it's not enough to bring an employment to the natural rate 00:39:13.550 --> 00:39:16.490 align:middle line:84% of unemployment to get inflation back to the 2% we like, 00:39:16.490 --> 00:39:19.190 align:middle line:84% because you need to bring expected inflation down now. 00:39:19.190 --> 00:39:23.690 align:middle line:84% And that means you need to bring the unemployment rate very, very 00:39:23.690 --> 00:39:25.980 align:middle line:84% high in order to re-anchor expectation. 00:39:25.980 --> 00:39:27.620 align:middle line:84% So that's a very scary situation. 00:39:27.620 --> 00:39:29.180 align:middle line:84% They were very persuasive, though, 00:39:29.180 --> 00:39:33.080 align:middle line:84% at the end of the summer with very hawkish speeches and so on. 00:39:33.080 --> 00:39:36.290 align:middle line:84% And they managed to re-anchor expected inflation. 00:39:36.290 --> 00:39:40.040 align:middle line:84% Expected inflation very quickly came down to 2%, 2.5% one year 00:39:40.040 --> 00:39:41.460 align:middle line:90% out even. 00:39:41.460 --> 00:39:44.330 align:middle line:84% But now it has been picking up again and now we're around 3% 00:39:44.330 --> 00:39:44.830 align:middle line:90% again. 00:39:44.830 --> 00:39:46.830 align:middle line:84% So it's a little scary where we are. 00:39:46.830 --> 00:39:50.050 align:middle line:84% So to me, this is going to be very important in that. 00:39:50.050 --> 00:39:53.830 align:middle line:84% So if inflation keeps lingering around 6% 00:39:53.830 --> 00:39:55.950 align:middle line:84% and so on, and eventually, expected 00:39:55.950 --> 00:39:57.990 align:middle line:84% inflation becomes unanchored, then there's 00:39:57.990 --> 00:40:01.880 align:middle line:84% almost no way around but to have a recession to get out of that. 00:40:01.880 --> 00:40:05.750 align:middle line:84% If that doesn't happen, if they succeed convincing people 00:40:05.750 --> 00:40:08.360 align:middle line:84% that they're very serious about this stuff 00:40:08.360 --> 00:40:11.300 align:middle line:84% and they re-anchor expected inflation, 00:40:11.300 --> 00:40:14.330 align:middle line:84% then we don't need to create a large recession. 00:40:14.330 --> 00:40:18.500 align:middle line:84% Still, they may create causes because accidents happen, 00:40:18.500 --> 00:40:20.330 align:middle line:90% but they don't need to. 00:40:20.330 --> 00:40:24.080 align:middle line:84% But they will need to if this guy gets unanchored. 00:40:24.080 --> 00:40:28.680 align:middle line:84% Actually, maybe I can use even this expression here 00:40:28.680 --> 00:40:30.150 align:middle line:84% to explain what I'm trying to say. 00:40:30.150 --> 00:40:32.640 align:middle line:84% And I realize that again, this is material really 00:40:32.640 --> 00:40:35.200 align:middle line:90% for the next lecture. 00:40:35.200 --> 00:40:37.020 align:middle line:84% What I'm trying to say is that if they 00:40:37.020 --> 00:40:45.220 align:middle line:84% manage to keep this theta very close to 0, then in order 00:40:45.220 --> 00:40:50.080 align:middle line:84% to bring inflation back to their target of pi bar, 2% or so, 00:40:50.080 --> 00:40:52.030 align:middle line:84% all that they really need to do is 00:40:52.030 --> 00:40:55.010 align:middle line:84% to bring unemployment to the natural rate of unemployment. 00:40:55.010 --> 00:41:01.730 align:middle line:84% So they only need to really fix this gap. 00:41:01.730 --> 00:41:05.150 align:middle line:84% They need to raise unemployment so it closes that gap. 00:41:05.150 --> 00:41:06.680 align:middle line:90% But it's a small change. 00:41:06.680 --> 00:41:12.430 align:middle line:84% That is, they succeed keeping expected inflation at around 2%. 00:41:12.430 --> 00:41:23.960 align:middle line:84% If they don't, suppose that theta becomes very far from 0. 00:41:23.960 --> 00:41:25.250 align:middle line:90% Then we have a problem. 00:41:25.250 --> 00:41:28.940 align:middle line:84% Because then expected inflation is above the target, no? 00:41:28.940 --> 00:41:30.050 align:middle line:90% Because we have 6%. 00:41:30.050 --> 00:41:31.430 align:middle line:90% So suppose that is equal to 1. 00:41:31.430 --> 00:41:35.610 align:middle line:84% We have 6%, then expected inflation is 6%. 00:41:35.610 --> 00:41:43.190 align:middle line:84% That means that if your expected inflation is 6%, then 00:41:43.190 --> 00:41:46.110 align:middle line:84% in order to bring the inflation-- if you bring 00:41:46.110 --> 00:41:48.360 align:middle line:84% unemployment just to the natural rate of unemployment, 00:41:48.360 --> 00:41:49.790 align:middle line:84% so the red line to the blue line, 00:41:49.790 --> 00:41:51.800 align:middle line:84% you haven't made a lot of progress. 00:41:51.800 --> 00:41:55.280 align:middle line:84% All that you have done is you have brought down inflation 00:41:55.280 --> 00:41:58.890 align:middle line:84% to 6%, which is expected inflation. 00:41:58.890 --> 00:42:01.670 align:middle line:84% So if you have expected inflation of 6%, 00:42:01.670 --> 00:42:04.340 align:middle line:84% you need to bring unemployment much higher 00:42:04.340 --> 00:42:06.110 align:middle line:84% than the natural rate of unemployment 00:42:06.110 --> 00:42:10.100 align:middle line:84% in order to bring inflation back to the target of 2%. 00:42:10.100 --> 00:42:15.260 align:middle line:84% That's the reason I say to me, the fight will be-- 00:42:15.260 --> 00:42:21.050 align:middle line:84% the battle will be won or lost on that term there. 00:42:21.050 --> 00:42:23.320 align:middle line:90% Yep? 00:42:23.320 --> 00:42:26.980 align:middle line:84% STUDENT: How much of this current inflationary pressure 00:42:26.980 --> 00:42:29.347 align:middle line:84% is caused by unemployment and how much of it 00:42:29.347 --> 00:42:30.430 align:middle line:90% caused on the supply side? 00:42:30.430 --> 00:42:33.260 align:middle line:84% Because it feels like a lot of the stuff, like CPI going up, 00:42:33.260 --> 00:42:35.860 align:middle line:84% energy prices going up, how much can the fed control something 00:42:35.860 --> 00:42:36.360 align:middle line:90% like that. 00:42:36.360 --> 00:42:40.180 align:middle line:84% RICARDO CABALLERO: Well, it varies at different places 00:42:40.180 --> 00:42:41.020 align:middle line:90% around the world. 00:42:41.020 --> 00:42:44.920 align:middle line:84% But in the US, for a while, a big component of the inflation 00:42:44.920 --> 00:42:49.430 align:middle line:84% was all that stuff, bottlenecks in the ports 00:42:49.430 --> 00:42:50.750 align:middle line:90% and stuff like that. 00:42:50.750 --> 00:42:52.965 align:middle line:90% That's almost all gone. 00:42:52.965 --> 00:42:54.340 align:middle line:84% There's very little of that left. 00:42:54.340 --> 00:42:57.030 align:middle line:90% So now it's aggregate demand. 00:42:57.030 --> 00:43:00.103 align:middle line:84% People feel very rich for a variety of reasons. 00:43:00.103 --> 00:43:02.520 align:middle line:84% They're spending a lot, and that's the reason unemployment 00:43:02.520 --> 00:43:04.100 align:middle line:90% is very low. 00:43:04.100 --> 00:43:05.710 align:middle line:84% It's not unemployment per se, it's 00:43:05.710 --> 00:43:08.920 align:middle line:84% just that aggregate demand is very high. 00:43:08.920 --> 00:43:11.470 align:middle line:84% And that translates into very low unemployment 00:43:11.470 --> 00:43:13.570 align:middle line:84% and that feeds into inflation this way, 00:43:13.570 --> 00:43:15.380 align:middle line:90% through wages and so on. 00:43:15.380 --> 00:43:19.300 align:middle line:84% But in the US, the component of aggregate demand 00:43:19.300 --> 00:43:20.650 align:middle line:90% is much larger than in Europe. 00:43:20.650 --> 00:43:25.510 align:middle line:84% In Europe, those supply side factors are much more important. 00:43:25.510 --> 00:43:35.080 align:middle line:84% So around yeah, the summer of 2022, you could say both Europe 00:43:35.080 --> 00:43:37.840 align:middle line:84% and the US are about the same amount of excess inflation. 00:43:37.840 --> 00:43:40.910 align:middle line:84% We're hold with around 10% inflation. 00:43:40.910 --> 00:43:44.950 align:middle line:84% But in the US was 2/3 excess aggregate demand, 00:43:44.950 --> 00:43:48.770 align:middle line:84% while in Europe was 2/3 problems on the supply side, 00:43:48.770 --> 00:43:53.020 align:middle line:84% especially because of the war and stuff like that. 00:43:53.020 --> 00:43:55.600 align:middle line:84% But for the US today, it's mostly an aggregate demand 00:43:55.600 --> 00:43:56.170 align:middle line:90% problem. 00:43:56.170 --> 00:43:57.910 align:middle line:84% We're not going to get a lot of-- 00:43:57.910 --> 00:44:00.860 align:middle line:84% obviously, if the war stops, that's going to help. 00:44:00.860 --> 00:44:02.510 align:middle line:90% But it's not going to be enough. 00:44:02.510 --> 00:44:05.710 align:middle line:84% We need to-- just the economy is too hot. 00:44:05.710 --> 00:44:07.510 align:middle line:84% It's too much aggregate demand out there. 00:44:07.510 --> 00:44:10.858 align:middle line:90% That's the fundamental problem. 00:44:10.858 --> 00:44:14.160 align:middle line:84% STUDENT: Can you explain again why an increase in z 00:44:14.160 --> 00:44:17.100 align:middle line:84% would increase the natural rate of unemployment? 00:44:17.100 --> 00:44:18.600 align:middle line:84% RICARDO CABALLERO: An increase in z? 00:44:18.600 --> 00:44:19.260 align:middle line:90% Yeah. 00:44:19.260 --> 00:44:26.430 align:middle line:84% So for that, the base is the previous slide diagram, 00:44:26.430 --> 00:44:27.960 align:middle line:90% but remember what z does. 00:44:27.960 --> 00:44:34.120 align:middle line:84% Actually, let me go to this equation here. 00:44:34.120 --> 00:44:37.170 align:middle line:84% So we can figure it out in these two equations here. 00:44:37.170 --> 00:44:42.870 align:middle line:84% If z goes up, that means for any given level of unemployment, 00:44:42.870 --> 00:44:46.650 align:middle line:84% unexpected inflation, wages go up. 00:44:46.650 --> 00:44:49.850 align:middle line:90% Workers demand a higher wage. 00:44:49.850 --> 00:44:53.425 align:middle line:90% But remember that the firms-- 00:44:53.425 --> 00:44:56.520 align:middle line:90% 00:44:56.520 --> 00:44:57.282 align:middle line:90% sorry. 00:44:57.282 --> 00:44:59.490 align:middle line:84% We're talking about the natural rate of unemployment, 00:44:59.490 --> 00:45:03.050 align:middle line:84% so let me replace this Pe for P, first of all. 00:45:03.050 --> 00:45:08.280 align:middle line:84% So I'm going to divide W by P on both sides. 00:45:08.280 --> 00:45:11.920 align:middle line:84% So I get-- if z goes up, the workers 00:45:11.920 --> 00:45:20.810 align:middle line:84% want a higher real wage, because if z goes up, then W over P, 00:45:20.810 --> 00:45:23.120 align:middle line:84% I'm dividing by P both sides, goes up. 00:45:23.120 --> 00:45:25.400 align:middle line:90% Workers demand a higher wage. 00:45:25.400 --> 00:45:27.320 align:middle line:84% But the firms, from here, you can 00:45:27.320 --> 00:45:30.620 align:middle line:84% see that I can divide by P both sides, W over P 00:45:30.620 --> 00:45:35.240 align:middle line:84% that the firms offer is equal to 1 over 1 plus m. 00:45:35.240 --> 00:45:39.110 align:middle line:84% So the firms are not going to offer a higher real wage. 00:45:39.110 --> 00:45:42.260 align:middle line:84% The workers want a higher real wage. 00:45:42.260 --> 00:45:44.330 align:middle line:84% The only thing that can restore equilibrium 00:45:44.330 --> 00:45:47.090 align:middle line:84% that the workers end up demanding the same real wage 00:45:47.090 --> 00:45:49.550 align:middle line:84% as the firms are willing to pay is 00:45:49.550 --> 00:45:52.730 align:middle line:84% that somehow, the hands of the worker gets weakened. 00:45:52.730 --> 00:45:55.970 align:middle line:84% And the only variable here that can weaken their hand 00:45:55.970 --> 00:45:59.120 align:middle line:90% is higher unemployment. 00:45:59.120 --> 00:46:01.910 align:middle line:90% So let me put it all in. 00:46:01.910 --> 00:46:05.480 align:middle line:90% 00:46:05.480 --> 00:46:10.070 align:middle line:84% So at the natural rate, I know that P is equal to P. 00:46:10.070 --> 00:46:19.010 align:middle line:84% So that means the wage setting equation implies W over P 00:46:19.010 --> 00:46:20.330 align:middle line:90% equals F u, z. 00:46:20.330 --> 00:46:22.970 align:middle line:90% 00:46:22.970 --> 00:46:32.080 align:middle line:84% From the price setting equation, I have that W over P 00:46:32.080 --> 00:46:34.930 align:middle line:90% is equal to 1 over 1 plus m. 00:46:34.930 --> 00:46:38.170 align:middle line:84% So in this very simple model, this is given. 00:46:38.170 --> 00:46:41.767 align:middle line:84% If this guy goes up, these guys want a higher real wage. 00:46:41.767 --> 00:46:44.350 align:middle line:84% But that cannot happen because that would be inconsistent with 00:46:44.350 --> 00:46:45.200 align:middle line:90% the price setting. 00:46:45.200 --> 00:46:47.920 align:middle line:84% So you need to bring down this guy. 00:46:47.920 --> 00:46:49.600 align:middle line:84% The only thing that can bring it down 00:46:49.600 --> 00:46:52.010 align:middle line:90% is for unemployment to go up. 00:46:52.010 --> 00:46:55.823 align:middle line:84% And that's-- at P, we call that a natural rate of unemployment. 00:46:55.823 --> 00:47:01.990 align:middle line:90% 00:47:01.990 --> 00:47:05.190 align:middle line:84% STUDENT: Last lecture, we talked about the labor force 00:47:05.190 --> 00:47:07.530 align:middle line:90% participation rate. 00:47:07.530 --> 00:47:11.362 align:middle line:84% Is there any reason to try and increase that to increase-- 00:47:11.362 --> 00:47:13.320 align:middle line:84% RICARDO CABALLERO: Oh, that would be fantastic. 00:47:13.320 --> 00:47:13.890 align:middle line:90% Yes. 00:47:13.890 --> 00:47:16.750 align:middle line:84% STUDENT: Is there a policy that [INAUDIBLE]?? 00:47:16.750 --> 00:47:19.240 align:middle line:84% RICARDO CABALLERO: Well, I mean there 00:47:19.240 --> 00:47:22.630 align:middle line:90% are negative policies as well. 00:47:22.630 --> 00:47:26.410 align:middle line:84% z reduction, in a sense, does that, because there 00:47:26.410 --> 00:47:28.750 align:middle line:84% was emergency unemployment benefits, 00:47:28.750 --> 00:47:32.330 align:middle line:84% an emergency income supplement and so on 00:47:32.330 --> 00:47:35.380 align:middle line:84% as a result of the pandemic that are disappearing slowly. 00:47:35.380 --> 00:47:37.150 align:middle line:84% And that's very natural, so it's going 00:47:37.150 --> 00:47:40.750 align:middle line:90% to bring participation back up. 00:47:40.750 --> 00:47:43.240 align:middle line:90% And it is beginning to pick up. 00:47:43.240 --> 00:47:46.900 align:middle line:84% So yeah, you need to incentivize return to work. 00:47:46.900 --> 00:47:49.450 align:middle line:84% And now there are some people that there is nothing, 00:47:49.450 --> 00:47:54.140 align:middle line:84% that they retire, essentially, or they have health problems 00:47:54.140 --> 00:47:55.840 align:middle line:90% and they just cannot return. 00:47:55.840 --> 00:47:56.920 align:middle line:90% We lost that. 00:47:56.920 --> 00:48:00.070 align:middle line:84% And the other margin, which is very important, is immigration. 00:48:00.070 --> 00:48:03.130 align:middle line:84% So that's a big issue, because immigration, obviously-- 00:48:03.130 --> 00:48:04.900 align:middle line:90% we lost, I think in the US-- 00:48:04.900 --> 00:48:07.720 align:middle line:84% I'm not a labor economist, but we lost, I think, 00:48:07.720 --> 00:48:13.480 align:middle line:84% a flow of the order of 500,000 people a year during COVID. 00:48:13.480 --> 00:48:17.850 align:middle line:84% And that's a big chunk of the decline in the labor. 00:48:17.850 --> 00:48:20.120 align:middle line:84% No, what you need is more employment. 00:48:20.120 --> 00:48:22.390 align:middle line:84% That's going to-- that puts downward pressure 00:48:22.390 --> 00:48:25.070 align:middle line:84% on wages for the same amount of aggregate demand. 00:48:25.070 --> 00:48:26.380 align:middle line:90% And that's what you need. 00:48:26.380 --> 00:48:29.810 align:middle line:84% But yeah, that's a very good point. 00:48:29.810 --> 00:48:31.540 align:middle line:84% We're taking all that as given here. 00:48:31.540 --> 00:48:33.910 align:middle line:90% Remember, we're fixing all that. 00:48:33.910 --> 00:48:37.210 align:middle line:84% But if you don't, then other terms 00:48:37.210 --> 00:48:40.680 align:middle line:84% start appearing in this expression and so on. 00:48:40.680 --> 00:48:42.405 align:middle line:90% Good. 00:48:42.405 --> 00:48:44.280 align:middle line:84% Obviously, I'm not going to start the review. 00:48:44.280 --> 00:48:45.630 align:middle line:90% We have only one minute. 00:48:45.630 --> 00:48:47.400 align:middle line:84% So in the next lecture, I'll just 00:48:47.400 --> 00:48:51.720 align:middle line:84% review the material for the quiz. 00:48:51.720 --> 00:49:04.000 align:middle line:90%