 I'm Susan Collins, the Joan and Sanford Wildein of the Gerald R. Ford School of Public Policy, and I'm delighted to see so many of you here with us this afternoon for another in our series of policy talks at the Ford School Lectures. It would help if I turned the microphone on as a good place to start. So again, welcome. I'm delighted to see so many of you here, and we have a very special program this afternoon. Our speaker is Christina Romer, who is the class of 1957 Garth B. Wilson Professor of Economics at the University of California, Berkeley. It's a special pleasure for me to welcome her here because we were contemporaries in graduate school, and so I have had the distinct pleasure of knowing Christie for quite a long time and have always enjoyed her insights and her thoughts. Well, initially, her work, she was known for her work on the causes of the Great Depression and policy responses and implications for policy today. Her recent research, however, has focused somewhat less on monetary policy and also on fiscal policies impact on economic activity. A hallmark of her work has been combining statistical evidence with a narrative approach as well that incorporates some of the much more nuanced information that we can get from looking at the historical record. While recognizing her expertise in the midst of the global financial crisis, she received a call from the White House, and from 2009 to 2010, she served as the chair of President Obama's Council of Economic Advisors. To list just a few of her honors, Christie is a Guggenheim Fellow, a Fellow of the American Academy of Arts and Sciences. She also serves on the NBER's Committee on the Dating of Business Cycles, which is perhaps particularly relevant in this context as well. And she is the recipient of the University of California, Berkeley's Distinguished Teaching Award. And I'm particularly grateful for that because we are one of the beneficiaries of your teaching, having recently hired Josh Hausmann to join our faculty, and we're delighted to have him here as well. Christie earned her BA in Economics from the College of William and Mary, and her PhD from MIT, where she met her husband, David Romer, who we're delighted is with us here today as well. David, welcome. Well, I'd like to remind our audience that if you have a question for Christine Romer, please write it on one of the cards that you should have received when you came into the room today. Ford School volunteers will pick up the cards at around 440 p.m., and they'll be circulating. So if you have questions, we look forward to them. We are delighted to also take questions from Twitter. And if you want to tweet a question in, please use the hashtag policy talks. During the Q&A session, we will have the questions read. So Josh Hausmann, his macroeconomic seminar, we have two students from that group who will be reading your questions, Yiran Chen and Ben Lusher. So with no further ado, it is my great pleasure to welcome Christine Romer to the podium. Well, thank you. It is such a pleasure and an honor to be here today. David and I were calculating, I think the last time I was here was my first year as an assistant professor. I'm gonna confess it was 27 years ago. So anyway, it is a delight to be back and to be here with you. So this afternoon, I'm gonna talk, as you can guess from the title up there, about what happens after financial crises and why. And kind of the issues the following, you know, back before 2008, modern economists kind of rarely thought about financial crises. Even though we were aware they could still happen, and indeed had happened in countries like Sweden and Japan in the not too distant past, financial crises or banking panics, as they're often called, were mainly relegated to the province of economic history. They were something that we could ignore because they were quite rare, and we thought fairly easily managed with our conventional tools of monetary and fiscal policy. But then we lived through 2008 and the meltdown on Wall Street. And I think we saw firsthand just how terrifying and how destructive a financial crisis could be. And that experience I think has renewed interest in panics and sort of what panics do to the economy and led to a whole rash of new research. And the conventional wisdom has just swung very much from the view we could ignore financial crises to the view that they are just incredibly important. Indeed, I'd say the new view is that crises give rise to recessions that are particularly severe and long lasting. And Carmen Reinhart and Ken Rogoff are the most well-known proponents of this view. Their book, This Time is Different, had the very good fortune to be published months after the failure of Lehman Brothers. And its view is that while policy makers always try to claim that this time is different, in fact, the aftermath of financial crises is usually the same and usually pretty wretched. And now I'd say just about everyone is adopting this view. So especially journalists and politicians on both sides of the Atlantic trying to explain why their economies are still struggling now five or six years after the financial crisis readily invoke the persistent negative effects of a financial crisis. And just recently Janet Yellen and her first press conference as our new Federal Reserve Chair in explaining why the US was likely to need easy monetary policy for quite a while yet invoked the long lasting devastation caused by the financial crisis. Well, in my talk this afternoon, I'm going to discuss some new evidence on this important topic. And it's actually, it's appropriate that David's here because it's based on a research that we're currently conducting together. And I think what you'll find is that our new research yields a somewhat more complicated picture of the impact of crises either than the old school, we can safely ignore them, or the new school, they are a death sentence for the economy. We find that crises certainly matter. They depress output in the near term, but their impact is on average pretty moderate and fairly short lived. Now, in addition to examining what happens on average following financial crises, we're also gonna look at the range of experiences. And here's I think the really important point. We find substantial variation in the impact of crises in different countries in different time periods. For example, Japan sank into a decades long slowdown following its financial turmoil in the 1990s whereas Sweden came just roaring back from its crisis around the same time. I think this range of experiences raises an obvious question, what accounts for these differences? And in my talk I'll try to suggest what I think are the most important factors explaining this variation. And then finally, the reason I think we care about understanding the range of experiences is very relevant to the audience here at the Ford School. And that is so that policy makers can figure out how better to respond the next time we face a financial crisis. So what does our analysis suggest might be effective strategies for dealing with financial disruption? And I'm gonna argue that measures to both prevent crises in the first place and policies to better manage their effects are important. And that this mix of proactive and reactive policies can help make sure that crises don't have severe and long lasting results. Or as I suggested in my title that the aftermath of a financial crisis just doesn't have to be that bad. Okay, so let me start on the impact of financial crises. What's the question of sort of what typically happens after a financial crisis? That is if we look over many countries and many time periods, what's the usual or the average result of a financial meltdown? And since Reinhardt and Rogoff's work is so well known I think it's useful to start by reminding you what they do and what they find. And actually their basic methodology is delightfully simple, right? So using a variety of sources, especially a study done by scholars at the International Monetary Fund, Reinhardt and Rogoff identify the dates of financial crises in a large sample of countries. And then they just look at what happened to real GDP per capita around the time of financial crises. And they're gonna focus on the percentage fall in GDP from its high to its low. All right, and here's a picture of sort of their key result. The names down the middle shows you a sample of countries and when they had a financial crisis according to their chronology. To the left you see the peak to trough decline in GDP and to the right the duration of that decline. So how many years was GDP falling? And the dark bar that I've highlighted and read there in the middle is the average. And what you're supposed to see is that the average fall in output per capita around the times of a financial crisis was about 9.3% and the average length of a fall was about two years. Now to put these numbers a little bit in perspective in a typical recession in the US since 1947, GDP only fell about 2% and the fall lasted significantly less than a year. So you can see that Reinhart and Rogoff's numbers do suggest sort of truly wretched economic performance following the average financial crisis. Well once we have this, an obvious question is, is there anything about what Reinhart and Rogoff do that should make us nervous about their findings? And here I actually want to be very clear, I'm just using Reinhart and Rogoff as an example. There are actually a number of studies that use quite similar approaches. And I think one drawback of this literature is that the definition of crises tends to be a little squishy. It has an element of we know one when we see one. Also the identification of crises is done after the fact by scholars who know what happened to output in the various countries. And we think that this combination of kind of fuzziness in the definition of a crisis and this sort of ex-post identification makes it possible for some bias to creep in. There's perhaps a natural tendency to look just a little bit harder for a financial crisis before a known severe recession. Finally, like the definition of crises, the empirical methodology used in this literature tends to be somewhat imprecise. Reinhart and Rogoff, for example, look just at the decline in GDP around crises. But in some places, the decline in GDP started well before the crisis and may have had nothing to do with the crisis itself. For example, one of the big observations here, Finland in 1991 is particularly interesting. Finland, one of its major trading partners was the Soviet Union. When we start to have the political upheaval and the eventual breakup of the Soviet Union, that delts a real blow to the Finnish economy. And so their GDP started falling and in fact, well before they had a financial crisis. And indeed, when we look at it, after their financial crisis, they actually grow just fine. So that most of that big fall really predates the crisis. All right, so those are some of the possible concerns or why we think there might be room for yet another paper in this literature. So David and I in our new research, you're gonna try to improve on existing studies in two ways. One is by identifying crises more accurately, we hope. And the other is by trying to use a more precise empirical technique to estimate their impacts. And I should point out this work is incredibly preliminary. I'd love to say that it's hot off the presses and the truth is it's not even that far along. It's literally hot out of the computer program. So it will certainly be involving and we hope improving as we actually get around to writing it up over the next couple of months. All right, so let me describe what we do to sort of what we think is to better identify crises since that's where most of the work in the project has been. To get around the problem of looking harder for crises when we know output went down in a country, David and I are gonna focus on the accounts of contemporary observers. And in particular, since 1969, an international group, the Organization for Economic Cooperation and Development, or the OECD as it's usually called, has written a long document every six months describing what's going on in a sample of advanced countries. And this volume called the OECD Economic Outlook appears to be very thorough and a fairly consistent in its reporting of what's going on in these countries, both across countries and over time. And our basic strategy actually picks up on something that Susan described was a common strategy of ours in research, is that we're going to use these sources. We're gonna read what the analysts of the OECD said about financial conditions in a country, and that that's how we're going to identify when crises or financial disruption more generally was happening. And our hope is that since these analysts are working in real time and they don't know what actually happened in the economy, that we're gonna get around some of those fears of bias that we mentioned before. Okay, so that's kind of, oops, I've been, I had all these nice slides. All right, so what do we actually do when we read these accounts? So we're gonna use real time sources, that's the OECD Economic Outlook. Then what are we gonna look for? What's kind of our definition of crises? Since I said there was sort of a fuzziness in existing definitions. So in particular, what we're gonna look at is what we call sort of descriptions of financial stress. And in particular, we're gonna look for evidence that contemporary observers felt that the provision of credit was being constricted by something other than tight monetary policy. Or more technically, what we're looking for is evidence that something was causing a rise in the internal cost of funds for banks relative to a safe interest rate. So things like balance sheet problems or a large number of loan defaults or just a rise in the perceived riskiness of financial institutions. And then an important innovation is that we're gonna scale the degree of financial stress from zero to 16. So rather than just saying, do we have a crisis or not, try to give you a range of what the level of financial stress is. Besides obviously giving us more information by having this scale of financial distress, we hope that it will also help to minimize errors. With the zero one variable, it matters a lot whether you call something a crisis. If you're allocating things along a scale from zero to 16, any mistakes that you make are gonna tend to be less consequential. Okay, so what's gonna come out of our analysis is a scaling of financial stress in 24 advanced countries. I should have said that. The OECD is about the countries of Western Europe, North America, Japan, Australia. And we're gonna do this every half year between 1967 and 2007. We're gonna stop before the recent episode, mainly because the recent episode's what we're trying to understand. And so we wanna use the past to give us a window into that. All right, so this graph shows you our new crisis variable when we see financial stress. It's showing you the 10 countries in the OECD for which we actually see some financial distress, a non-zero observation. The other 14, the analysts at the OECD never see any financial stress. Now you can see a number of things. Here one is there basically was no financial distress anywhere in the 70s and the 80s, but in the 90s you actually have a fair amount in a number of countries. Another thing that certainly is true is that in our measure we see most of the same things that show up in other chronologies. So for example, if you can see it, the purple here, that's Sweden, right? They had a famous financial crisis in the early 1990s. The black line that just goes up and stays up, that's Japan that indeed had a lot of financial distress for a very extended period of time. Now though the same things kinda show up in our measure, most of the same things as in other chronologies, the timing is actually often quite different. And so this is just an example that's a little bit close to home. This is the United States in the SNL crisis. Those two blue lines are Reinhart and Rogoff's start and end date for when they say there was a crisis in the U.S. And then the red line is our measure of financial stress and you'll see that the analysts at the OECD absolutely said there was a period of pretty acute financial stress here in the U.S., but it's much later than sort of the 1984 that Reinhart and Rogoff choose. They don't see it at the OECD until really 1990. And in general, if there's sort of a common thread here is when we date crises using our real time sources, they're almost always a fair amount later than most of the existing chronologies. Okay, so that's sort of what comes out of our, the sort of the input that we're going to generate. So now the whole question is, well, what does our new series tell us about, on average, what happens to output or the economy after a financial crisis? And here's where our second, what we like to think of as an innovation is going to come in. We're going to try to have what we hope is a more precise empirical approach. And our methodology is pretty straightforward. We're going to have data from the OECD countries on the two we're going to focus on, our industrial production. So the output of mining and manufacturing and then real GDP. So a very broad indicator of economic conditions. And what we want to know is sort of what happened after financial crises. The way we're going to do this is to run regressions. We're going to, there are going to be time series cross-section regressions, by which we just mean we're going to use the variation output in crises across countries and across time to identify the impact of crises at different horizons. Do countries having crises do worse than countries that don't? And in this process, we're going to control for lots of things. So lagged output, country specific factors, precisely so that we can take into account what was normally happening in each country. And then finally, we're going to summarize the results with an impulse response function. And that just means it's going to show us what happens to output for the five years following a crisis relative to what otherwise would have happened. And to give you a sense of magnitude, so you'll be able to kind of scale them in your head, when we do this simulation of what does a crisis seem to do to output, we're going to consider a fairly big financial disruption. So think of our scale from zero to 16, we're going to look at an eight, so sort of halfway in between, but a fair amount of disruption. Okay, so let me start with the results for industrial production. So this picture just shows you what's going to happen to industrial production following a moderate crisis if you look at our full sample of 24 countries. And the solid line is the point estimate of the impact and that dashed line is the two standard error bands. So to give you a sense of your sort of statistical precision. And what you're supposed to see from this is in, and this is in half years after the crisis, in period zero, so the same half year as you have the crisis, industrial production falls 4%. And it's highly statistically significant, that's why those bands are fairly narrow. Now important thing is that's actually not that big and a way to kind of get a sense of that and work David and I did probably 27 years ago and presented here when we were here last, was looking at the effects of pretty extreme, we have a period of monetary contraction in the US, a kind of run of the mill, the Fed decides it wants to get inflation down. We found that those kind of episodes reduced industrial production about 12%, right? So about three times as much as we estimate a financial crisis does. The other thing that we find pretty striking is just how quickly the results go away. Remember the fact we thought we learned from Reinhard Rogoff is that the recessions after a financial crisis are not only severe, they're really long lasting. And this is saying that by two years after the crisis, you're certainly back to normal and actually well before that, you're recovering quite significantly. All right, now here's the results for real GDP. So as I mentioned, that's a broader indicator. Again, let's look at that contemporaneous impact when you have a crisis in our sample of countries, it looks like real GDP falls about three, three and a half percent. Again, and it's highly statistically significant. So that's why we always wanna say there absolutely is an effect there. But again, I wanna give you the sense it doesn't seem to be that big. And again, remember Reinhard Rogoff's number was 9.3%, right? So again, about three times what we're finding. Now for GDP and the full sample, oops, oh, sorry. I'm familiar, clicker. Oh, wow, we really went through a lot. Okay, there we go, nope, one more. For the full sample, it looks like the effects are pretty persistent. The GDP falls and then actually stays low for the next five years, though the standard errors are starting to get a little fatter on you. Interestingly, that's being driven very much by the observation for Japan. So I already mentioned Japan had a lot of financial crises. They also had some uniquely bad GDP performance. And here's just kind of to help you focus in on this. This is our crisis variable for Japan. And what you're supposed to see is unlike many countries, when they have a financial crisis, it doesn't like go away. It's there for more than a decade. I think this is like 15 years, including two periods where it goes really high. That's the highest measure we get in any country is up here. And this one is also certainly over what we think of as that kind of eight pretty significant financial distress. The reason I emphasize this is if you take Japan out of the sample, the results look very much like for industrial production. So you get a fall in GDP again, about 3%, but it goes away actually again within sort of that two year period. So it depends on whether you think Japan is a very telling example and you wanna put a lot of weight on it or do you think it's kind of weird in which case you might wanna leave it out of your sample. The important thing is it matters. So that's our finding. All right, so sort of where does all of this leave us? This is kind of our new empirical work. I think what I'd like you to see from this is that the simple conclusion that crises are obviously horrible for an economy I think just really doesn't hold. I think our bottom line is it's more complicated than that. So we absolutely think we see a strong near term negative correlation between crises and output but it's just not very big and except in the case of Japan the effects don't seem to last for very long. Okay, so that's kind of the empirical work. I think this reference we've already gone down the road of talking about Japan that kind of leads very naturally to the next topic that I wanna discuss and that is the variation in the impact of crises and possible explanations for it. So again, kind of think about what we've just done. So the impulse response functions that I showed you really captures sort of the usual or the average response output to crises in all of our episodes. But we think there's actually a kind of straightforward way to say, well, can we get a sense of the variation in what happens to output after crises? And the way we're gonna do this is actually to let's start with just a sample of the periods of kind of extreme financial stress that we find. So all of the episodes that we have where our measure goes above an eight and we actually have six of those or seven if you count the fact that Japan hits eight twice. And what we're gonna do is kind of a thought experiment. We're gonna do the very simple procedure of let's just make a simple forecast of where output was headed before the crisis. So just using lagged output and country specific factors, what would you have predicted was gonna happen to the crisis up through using data up through the half year before the crisis hits. And then just compare that kind of simple forecast to what actually happened. These are the six cases where we have a significant amount of financial stress. The red line in each case is that simple forecast of where the economy was headed. And the blue line is what actually happened. And what I hope you will see is that and for the cases, so the US and all of these are in the early 90s, Norway, Sweden and Finland, the actual line in that simple forecast aren't really very different. And that's kind of a way of saying what happened to output was actually pretty well captured by the trajectory that we were on right before the crisis. So it doesn't suggest much of kind of an additional impact of the crisis itself. But for two countries, Japan and Turkey, it sure looks like that actual line is way different from your simple forecast. And that's a way of saying at least in those two episodes, it does look like the aftermath of a financial crisis was truly wretched relative to what you would have forecast or what it looked like, the trajectory that they were on, right? So that's a way of capturing very visually the amount of variation we see in the output consequences of a crisis. All right, well, that just kind of pushes the mystery back a bit. We know there's variation. What might explain it? And I think perhaps the most important factor that I'd mentioned is just persistent or repeated crises. Perhaps the most important reason that some cases are so much worse than others is that the crises themselves are so much worse than others. There's so much more persistent and severe. And here I can't resist a bit of a digression on the Great Depression here in the United States because you know the depression is often given kind of as the poster child for the idea that financial crises do indeed have long lasting effects, right? So in the usual telling of the story of the Depression, we had a financial crisis in 1929 but the next three years output plummeted and for the seven years after that, output was still pretty lousy, right? So that certainly sounds like a financial crisis having a very severe and long lasting consequence. Now the first problem with this conventional story is that the US didn't actually have a financial crisis, at least as we would define it in 1929. There was indeed a big fall in the stock market but this is actually one time when the Fed actually stepped in and helped destabilize the financial system. And the result was that there were actually relatively few bank failures and not much disruption in the financial system, not much of a rise if you want in the internal cost of funds for banks, at least for the first year or so. And this picture really helps you to see that. So this is a picture of the deposits in failed or suspended banks, right? So just how many banks were getting into trouble in the late 1920s and the early 1930s? And what you're supposed to see is not very many in 1929 and 1930 that the first big wave of bank failures and real distress in the financial markets doesn't occur until October of 1930. And in fact, this is, I've just put in vertical lines where Friedman and Schwartz, the very famous historians of the Great Depression, identify four waves of banking crises. So October of 1930, two in 1931, the spring and the fall, and then the last one was early in 1933, just leading up to Roosevelt's inauguration. And then if this is a picture that shows you industrial production, a measure of output, and then the vertical lines are those same Friedman and Schwartz, when did we have financial distress? What you're supposed to kind of see is it does indeed look like a big cause of declines in output are these waves of banking crises that after each, at least to the first three, output does indeed go down following these waves of banking crises. And that kind of makes sense. They destroy confidence. They make it hard to get credit, all the things that we usually think financial crises might do. Something else though that stands out from this picture is what the heck happened after the last wave of banking crises, right? So Roosevelt, as I said, came in in the midst of what's arguably the worst of the banking crises, and yet the very next month industrial production started to rise. So I think that far from showing that the effects of financial crises are inevitably terrible, maybe what the depression provides is some of the most compelling evidence that in fact a financial crisis doesn't have to devastate the economy for a long time afterwards. And I'll say a bit more about why output recovered so quickly in the spring of 1933 in a little bit. The important thing that I want you to take from this is why was the depression so horrible? It wasn't because we had one financial crisis that then had these severe effects. It was that we had these waves of crises that just wouldn't go away. And I think the analogy to Japan is really very clear, that a key reason I think why Japan's experience was so awful is that its financial crises themselves itself was incredibly persistent. And I showed you that picture before. Japan had some financial disruption for more than a decade. And so I think the bottom line from the experience of both the depression in Japan is that one of the reasons why the aftermath of crises is sometimes so bad is that the crises themselves are sometimes so bad. All right, factor number two, having a currency crisis at the same time. So as I've described, a financial crisis involves a disruption in the supply of credit. So banks fail, credit spreads rise, credit becomes hard to get. A currency crisis is something different. So typically this affects a country that has a fixed exchange rate. For example, Turkey in 2001 had pegged its exchange rate to the dollar and the euro. And in a currency crisis, people lose confidence that the central bank and the rest of the government will be able to maintain that fixed exchange rate. And usually what ends up happening is the exchange rate falls precipitously. And previous work has found that financial crises that happen at the same time as a currency crisis seem to be followed by more severe declines in output. Now why this is, maybe it's just having two of these things that once has a bigger effect on confidence, a bigger effect on wealth. But it might also be in terms of the policies used to deal with the currency crisis. What does the central bank do confronted with a loss of confidence in its currency and its exchange rate, it often jacks up interest rates which may deal a blow all of its own to output in the rest of the economy. Well the reason I mentioned this is remember Turkey was our other case where the output consequences of a crisis look particularly bad. And one possibility is what was going on in Turkey is they also had a currency crisis. And you just see, this is just a picture of their exchange rate. It did indeed plummet about the same time that they had a financial crisis. Okay, so that's factor number two, factor number three, other shocks, right? And this is actually I think a really important idea. You know, crises don't in fact usually come out of nowhere. For the most part, they're often triggered by something and whatever triggers the crisis often has effects of its own on the economy. And the most obvious example of this is an asset price bust. So think about 2008. Why did our financial markets get into so much trouble? It was the big run up in house prices, all of the sort of overbind and bad lending that went with that. When house prices came down, financial institutions got into trouble. But when house prices came down, that had also had sort of effects on the economy above and beyond anything having to do with the financial crisis, right? So we know that a lot of homeowners suddenly found themselves, well, with a lot less wealth because their house had gone down in value, a lot more debt relative to their wealth and that that may have affected their spending. This actually is just a nice picture from a recent study. It's a scatter plot. So along the horizontal axis, we have what was the change in house prices. So as you move to the left, house prices are falling more. And on the vertical axis, this is the change in consumer spending. And each one of those dots is a county in the United States. And so what you're supposed to see from this picture is that counties that had more of a drop in the value of houses also had a bigger drop in consumer spending, right? So that's a way of saying, above and beyond any effect of the crisis, there was an effect on consumer spending and output coming from just the collapse of house prices themselves. So I think one of the reasons if you wanna say, why was the aftermath of the 2008 crisis so horrible, it's probably not just that the financial crisis was bad and caused problems. There were other shocks that were also bringing down the economy at the same time. All right, the final factor I wanna mention in explaining this variation in outcomes across episodes is the policy response. I already mentioned that one reason the combinations they have a financial crisis and a currency crisis may be so bad involves policy. The actions taken to defend the currency like a rise in interest rates may make the impact of the crisis worse. And again, I can't resist coming back to the Great Depression because this is another one of those episodes where policy clearly played a role in exacerbating the decline in output. Remember, as I discussed, we had these waves of banking panics in the early 1930s. So obviously the Federal Reserve didn't take aggressive enough actions to really stop these waves of panics. But I think something that is less well known and even more surprising is that actually in the midst of all of this and in the midst of unemployment going up to 20 or 25%, they actually took some major contractionary policy actions. And in particular, in 1931. So this is a picture of the discount rate, which is just the rate at which the Federal Reserve lends to banks. And one of the things you see, remember I said that when the stock market crash happened in the fall of 1929, the Fed kind of tried to help, right? So they dropped the discount rate and that's part of why we don't see financial panics immediately or certainly banking panics immediately in 1929. But what's really striking is in October 1931, the Fed decided to raise the discount rate by two percentage points. And they did this because in September 1931, Britain decided to go off the gold standard. And so to reassure investors that the United States wasn't gonna go off the gold standard, they said, we'll show you, we'll raise our interest rate and stab ourselves in the foot. But anyway, but that, and that was a pretty consequential action. So in essence, they were taking this measure to try to stave off a currency crisis. Similarly, on the fiscal side, when you have a terrible decline in output, your budget deficit tends to go up because tax revenues are down. And Herbert Hoover in 1932 was very concerned by the skyrocketing budget deficit in the US. And so he decided to have a significant tax increase, again in the middle of a very severe economic downturn. And this mix of contractionary monetary policy and contractionary fiscal policy is yet another reason why the depression was so wretched. Well, I think something similar, at least on the fiscal side, surely explains why the 2008 financial crisis has seemed to have such severe and persistent effects, particularly in Europe. Following the meltdown in Greece in the spring of 2010, investors lost confidence in the bonds of a number of the Southern European countries. So countries like Spain and Portugal came under extreme pressure to raise taxes and cut government spending. And I actually remember vividly taking a trip to Europe in May of 2010 with then Secretary of the Treasury, Tim Geithner, to actually try to urge some moderation in how much a fiscal restraint was happening. Now, in truth, the thing I remember best about that trip was the plane flight. So I'll tell you a little bit. We were flying in a military plane and even though both of us had cabinet rank, the chair of the council of economic advisors, the secretary of the treasury, but one thing you quickly learn in Washington is that the secretary of the treasury completely outranks the CEA chair. And what that meant in this particular case was that Tim got the one state room on the plane. And at some point in the middle of the night, we hit some pretty extreme turbulence and I saw Tim's security detail rush to his cabin. And being a quite nervous flyer, I grabbed one of the agents like, are we going down? And he said, no, ma'am, we were just afraid the secretary might have fallen out of bed. So I allowed us to how we should all have such worries. But the message that we were trying to deliver on that trip is that if everybody rushed toward fiscal contraction at the same time in Europe, it could be devastating. Unfortunately, that message did not seem to have gotten through and indeed many of the countries undertook extreme fiscal contraction and we see in countries like Greece and Spain that have had to take sort of some of the most severe fiscal austerity measures. Their unemployment rates are now up in the low 20% range. The reason why I mention this and spend some time on it is I think fiscal contraction is an important part of why the recovery in this particular episode has been so slow. It's not that financial crises always lead to large and long-lasting falls in output. Rather, contractionary policy dealt a second serious negative shock to many economies just as they were starting to recover. All right, so the bottom line of this discussion is that I think several factors explain why the aftermath of crises is sometimes so much worse than others. Particularly persistent or repeated financial crises tend to lead to worse outcomes. Combining a currency crisis and a financial crisis can make things worse. But it's also the case that other shocks including contractionary policy may account for some of the variation that we see across episodes. Okay, so so far I've talked about sort of what typically happens and what explains some of the variation in what happens after crises. The last topic I wanna talk about is in some ways the most important, which is, well, what does all of this analysis suggest about possible strategies for dealing with crises? What can policymakers do to try to ensure that the aftermath of financial crises isn't that bad? And here I'd say surely the first and best piece of advice is don't have a crisis, right? So our new evidence suggests that financial crises are not the knockout blow, perhaps that the other estimates suggest they might be. But even in our new chronology and our new analysis financial crises are definitely a shock that takes a toll on the economy, on output, on employment, particularly for the first six months after a crisis. And I think they have the potential to be very painful if the crisis is allowed to drag on or if it's compounded by other factors. So surely the best thing don't ever start down that road, all right, so that just again kind of pushes the mystery one level back. What could we do to not have financial crises and we could have another whole several hour talk on that topic. But I'll mention what I think is the most important, which I actually think is better financial regulation. In particular, I have to say I've been persuaded by some of the recent research that suggests that sort of the simple, the most effective way to ensure financial stability is to just have higher capital requirements for financial institutions. That just means that banks and anything that operates like a bank would be required to have a larger amount of invested capital. I think requiring financial institutions to have a bigger cushion of invested capital does at least two things. One is it means owners have a lot of skin in the game and so they're naturally gonna act more responsibly to protect their investment. And it means that there's more invested capital to take losses if there's a severe downturn and loan defaults rise. And as a result, depositors and other creditors know the bank is likely to remain solvent so they're not gonna panic when trouble starts to happen. Okay, so I think clearly preventing crises, that's the first and best strategy. So what do you do if a country forgot to follow rule number one? What's the best crisis if it was the best strategy if a crisis nevertheless happens? Well, strategy number two, if it happens, get it over with as fast as possible. Right, and that kind of, I think that makes a lot of sense. One of the things that came out so strongly from say what I showed you about Japan or the Great Depression is that one of the main things that makes the outcome of a crisis much worse is if it drags on or if you have repeated, persistent financial crises. So to prevent that, governments need to move quickly to stabilize the financial system. And sometimes you can do that just sort of the old fashioned way, provide a lot of liquidity. This is the quintessential, what's a central bank supposed to do in a crisis? It's supposed to just lend any bank that wants cash, cash, as long as they have good collateral. And certainly the Federal Reserve did that in a big way in 2008. Unfortunately, it often takes a lot more than that to stop a modern panic. And indeed, I think one of the things we've seen is that governments often may have no choice but to actually bail out financial institutions. They may need to absorb some of the losses and rebuild the capital buffers before the panic actually subsides. And this is really, this is what the TARP legislation did back in 2008 and 2009. So if you're not up on all your acronyms, right? So this was a piece of legislation with the blessing of President Bush. So it was done before President Obama came into office. So it was Secretary of the Treasury Hank Paulson, Fed Chair Ben Bernanke, got together and convinced Congress to approve $800 billion of funds to stabilize the financial system. And the main things that those funds were used for was to just build up the capital reserves of our biggest financial institutions. And I can tell you as someone who had to interact a lot with Congress in January of 2009 and December of 2008, this is probably the single most hated piece of legislation ever, both by Congress and by the American people. But I think importantly, it worked, right? That it played actually, I think a crucial role in stabilizing the American financial system. And because it worked, in fact, banks were fairly quickly able to raise private capital to strengthen their position and the government ended up getting most of that $800 billion back, all right? In addition to stabilizing the financial system, governments often need to do even more than that to make banks actually healthy and ready to lend again. And so one of the things that I think when people say, what went wrong in Japan? It's in fact, the government did just enough often to stop the most acute part of the crisis, but not enough to actually make their financial system healthy, right? So for much of the 1990s, what we often say is that Japanese banks were zombies, so they were still operating, so they were technically living, but they were basically insolvent, right? So they were dead in the economic sense. And they had the basic problem. They had a bunch of questionable loans on their balance sheets, and so they were afraid to do more lending. And this made them prone to further ways of panics. It was really only after the Japanese government sort of made banks in Japan realize their bad loans and put in a significant amount of public capital did their financial system actually start working again. Now the classic sort of counter example to Japan is Sweden, right? So Sweden had its financial crisis in 1993, and in that case, the government moved incredibly quickly. So you can actually see this is our measure of financial stress in Sweden, and what you see is it absolutely shoots way up to a pretty significant level, but then it comes back down within a year. So just think in your mind how different the picture for Japan is, where it stretches up and elevated for almost 15 years. In Sweden, what happened is the most troubled banks were kind of immediately nationalized. Their losses were immediately realized. They were recapitalized by the government, and then they were sold back to the private sector. And I think experts think that this very aggressive approach was a key reason why the panic was such a sort of one-off event, and why Swedish growth recovered so quickly. All right, so strategy number three, deal with accompanying shock. So in some sense, everything I've talked about so far has to do with the crises themselves. So either don't have one or if you have one, get it over with as soon as possible. But one of the things or what else might be helpful, I think one of the things I've come to feel is important is to actually deal with some of the other shocks that may be going along with the financial crisis. So again, kind of come back to the case of the United States in 2008 and 2009. As I've described, this collapse of house prices obviously wreaked havoc in our financial system, but it also wreaked havoc on the balance sheets of ordinary families. That we found ourselves with a lot less wealth, we had taken on all this debt in the mid-2000s. And so a lot of households just really were afraid to spend, right? They were trying to pay down their credit cards, their home equity loans, and that kind of things. And so they weren't out there buying cars and appliances and all those things that employ people when we produce cars and appliances and things like that. And actually this actually resonates with me because I'll confess that back in 2009, I was actually somewhat skeptical of measures sort of to deal directly with the debt loads of homeowners. I wrestled with this, but it seemed to me that the people who owned homes were already sort of in the upper half of the income distribution. So when we thought about doing programs that might target help particularly to homeowners, that made me uneasy to simply because they were better off than many of the other people that were suffering during the crisis. But I think, no, if I knew then what I know now about sort of how slow the recovery was gonna be, and in particular how slow consumer spending was gonna be in coming back precisely because of the effects of these high debt loads, I think I might have proposed or certainly worked for other policies. And certainly one of the things I really regret is I think I would have pushed much harder had I again kind of known what was gonna happen for changes in the bankruptcy law. So I think one of the key facts about our current bankruptcy system is that a bankruptcy judge can't write down or restructure a home loan. So the only recourse there is to put it into foreclosure. And I think making it possible for a bankruptcy judge to write down the principle on a loan, a move colorfully referred to as crammed down, I think makes a lot of sense, right? Would have gotten some of the most troubled homeowners out from under those very high debt burdens and actually ended up probably lowering the losses for banks because a lot of value is lost when a home has to go through foreclosure. But sort of as a more general matter, I think policymakers need to look at what else is going on in the economy along with a financial crisis. And often they need to deal directly with those other shocks. All right, strategy number four. Well, use monetary and fiscal policy aggressively, right? So this four strategy just says we've got some conventional tools, fiscal policy, monetary policy. They're potent, they're effective and faced with a financial crisis, you should use them very, very aggressively. I think this is a lesson that was followed somewhat in 2008 and 2009, but not as much as it needed to be. So if you again kind of remember back in 2008 and 2009, the Fed absolutely took some very expansionary monetary policy. They quickly dropped their interest rate that they control, the federal funds rate to zero. And then they took even some actions beyond that, like buying a whole lot of mortgage backed securities to try to push down mortgage interest rates which weren't already down to zero. Likewise, the US did a very large fiscal stimulus, the American Recovery and Reinvestment Act. And I have to tell you that though like the TARP, the Recovery Act has become very controversial, I think it was incredibly helpful. And importantly, that's not just my opinion. It's in fact the conclusion of the nonpartisan Congressional Budget Office and about a half dozen recently published research studies on just what did the Recovery Act do. But if you were to ask me what was sort of the main mistake that both the Fed and the administration made, it was to not be even bolder in our policy response. I think the actions taken were absolutely helpful. They explain a lot about why the US economy turned the corner in the summer of 2009. But I think we all know that that recovery has not been nearly as strong or as thorough as any of us would have liked. And so one lesson that I take from the 2008 experience is that in responding to a financial crisis, policy needs to be very, very aggressive. So extraordinary times really do call for extraordinary responses. And again here, I'm gonna keep coming back to the Great Depression because probably the best example of a truly bold and aggressive monetary and fiscal response was Franklin Roosevelt. And in fact it was such a bold response as often referred to as Roosevelt's regime shift because he really fundamentally changed the policy regime. What did he do? Well first he came in in the middle of a financial crisis to try to stop that. He basically shut the entire financial system. We had a bank holiday that basically said you couldn't reopen your bank until someone had been in to inspect the books and declared that it was safe. And this is something that seems to have got a, given the economy or the American people a sort of cooling off period and it seems to have stopped the bank runs. Probably even more extraordinary were Roosevelt's monetary actions. So like 2008 by 1933 interest rates were basically down to zero so we didn't have our conventional monetary policy tools. But Roosevelt nevertheless found a way to do a big monetary expansion. Mainly he took us off the gold standard roughly a month after coming into office. We were off the gold standard. He allowed the dollar to depreciate about 30%. We had a big increase in the golds. The lot of gold flowed in. He used that to increase the money supplies. We actually had a quite significant monetary expansion. And then as Josh's wonderful work in his dissertation talks about there was a big fiscal expansion as well. So starting as early as 1933 the government budget deficit certainly went up because we passed all sorts of relief programs, public works, things like that, veterans bonuses. The important thing is this sort of regime shift absolutely had an impact. And it had an impact that a modern policy maker can only dream of, right? So this is a picture of the red line is the unemployment rate. The blue line is what happened to real GDP. That gray vertical line is when Roosevelt came into office at the start of 33. And just look real GDP increased 11% in 19, between 1933 and 34. The unemployment rate fell three percentage points in one year, right? So just an incredible amount of recovery very, very quickly. I think the best modern example of a truly bold policy response probably comes outside the United States in 2009. And in fact, in a number of Asian countries confronted with financial difficulties and plunging trade, especially China and South Korea undertook fiscal stimulus programs that were significantly larger than our Recovery Act, at least relative to the size of their economies. And the growth in those economies recovered substantially more relative to what had been happening before than in countries taking smaller actions. And one reason I think this is such an important point to make is that one of the reasons why China and South Korea could be so bold in their fiscal policy response in 2009 was that they started the recession with very low levels of government debt. So I think one corollary to the lesson that policy needs to be bold in response to financial crises is it's actually really important to run responsible fiscal policy in good times because you never know when you might need to run irresponsible fiscal policy in bad times. All right, last strategy, avoid self-inflicted wounds. So in some sense, all of my previous advice have been do this, do this. This one is don't do that. And the don't is, as I said, cause self-inflicted wounds. And by this what I mean is policy makers need to refrain from actions like monetary and fiscal contraction that make outcomes worse. Now you might think this is a pretty obvious piece of advice and it wouldn't even warrant taking a few minutes to say, but if that didn't one common feature of financial crises is that countries do seem to compound the problems by switching to contractionary policy prematurely. And perhaps the clearest example of this of a self-inflicted wound is the United Kingdom in 2010. So unlike, say, Greece and Spain and some of the countries of Southern Europe, Britain wasn't actually under any pressure for markets to get their budget deficit down. Nevertheless, in 2010 the new British government decided now would be a great time to do a fiscal austerity program. And if you look at, this is a picture I just love, it shows real GDP, the red line is the UK, the blue line is the Eurozone and the yellow line's the US. What you're supposed to see is kind of everybody was starting to grow coming out of mid-2009. And then Britain, when they did their fiscal austerity program, basically put themselves back into recession and a period of really roughly three years of almost no GDP growth. And it's only really been in kind of the last year that GDP has started to grow again in the UK. The same thing tends to happen with monetary policy. One of the most frustrating things for a monetary economist is to look at what happened in Japan following their financial crises in the 1990s and the 2000s. At one point they were actually starting to get some traction doing some expansionary monetary policy and they stopped sort of just as it was beginning to pay dividends. And importantly, it's not hard to understand why this premature switch to contractionary policy frequently happens. Financial crises tend to cause governments to have to spend a lot of money, either to bail out their financial system or to do a big fiscal stimulus. And policy makers become nervous when they see rising levels of government debt. So a natural response is to say, let's get that budget deficit down as quickly as we can. Likewise, central bankers confronted with a financial crisis often have to flood the system with liquidity so the money supply gets big. And before long monetary policy makers start to get nervous that this swollen money supply is gonna cause inflation. So they take measures, let's get that down as soon as we can. I think more than anything, I suspect that policy makers like everyone else following a traumatic event just want to get things back to normal. And I think there's a tendency to think that if they normalize policy that will help to accomplish getting back to normal. But in fact, that logic is almost always wrong that by withdrawing policy support from an economy too soon premature policy tightening actually prolongs the pain of the crisis rather than hastens the recovery. All right, so we've obviously been through through lots of material this afternoon. Very quickly sort of where does it all leave us? I think my bottom line is financial crises obviously matter. They are a shock to the nervous system of the economy and even our new estimates as I suggest say that crises take a toll on the economy at least in the near term. I think my whole point though this afternoon is that a crisis doesn't have to be a mortal blow. Some economies actually do reasonably well following a financial crisis. And therefore we shouldn't just throw up our hands and say we're all doomed if a financial crisis occurs. I think better policy along many dimensions can just make a huge difference. Ending the crisis quickly, taking aggressive actions to restore growth. Those are just some of the keys to making sure that the aftermath of a crisis isn't that bad. And as I've tried to suggest this afternoon I think kind of no episode kind of captures everything that I've been saying better than the Great Depression. Though we often think of the Depression as one crisis and one terrible downturn as I've described it was in fact a very complex event with waves of banking panics or crises. There were banking panics in October of 1930 and 1931 which were indeed followed by plunging output and skyrocketing unemployment. But there was another wave of panics as I've suggested probably the worst in early 1933 that was in fact followed by rapid growth and robust recovery. So what explains the difference? Obviously lots of things mattered but I think a really important one was policy. Herbert Hoover failed to stop the crisis and compounded their effects with contractionary monetary and fiscal policy. Franklin Roosevelt on the other hand moved quickly to stabilize the American financial system and took aggressive monetary and fiscal actions to try to get growth going again. So I think we need to learn from this example and all of the others that I've talked about this afternoon what happens after a financial crisis isn't written in stone. It's largely in the hands of policy makers and we need them to be Roosevelt's and not Hoover's. Thank you. I am a master's student here and I'm a for-to-over policy. So our first question today comes from the audience. How much does the zero lower bound on nominal interest rates matter to the aftermath of financial crisis? If nominal interest rates are at zero is the aftermath worse? That's a great question. Last night in the hotel room I said, oh shoot, you know I forgot to mention the zero lower bound. So anyway, so you're just confirming all of my fears. So actually the reason I didn't mention is I couldn't figure out where to put it in the talk. So it's about the policy response which was the last part and yet as you described it may be a factor that explains why some aftermaths are worse than others. So and I think that's the right way to think about it. So for anyone that isn't up in all of this, so what would the Fed normally do and confronted with a fallen output or a crisis or whatever is gonna drop interest rates? And as I described in 2008, the Fed absolutely did that and by December of 2008 they dropped the funds rate effectively to zero, right? And that's what we mean by the zero lower bound. And I think it absolutely is a part of what has made this episode particularly bad is because we didn't have sort of our usual tool of monetary policy that we could, you know, the quick and easy let's just keep dropping interest rates. Likewise, if you think about what some of what went wrong in Japan, you know, probably for the first decade they didn't hit the zero lower bound but for the second decade that they had financial crises, they were at the zero lower bound. So that is one of the things that probably constrained the policy response. So I do think it is something that matters. Part of why, you know, I think the case of say of Roosevelt matters is we were at the zero lower bound then as well. And Roosevelt nevertheless found that there are other monetary actions you can take, just swelling the money supply, the sort of something very close to what we'd call quantitative easing today. And it actually worked really well. So even though you were at the zero lower bound it was a tool that was still there. So it certainly would be, I think the aftermaths are gonna tend to be worse when you're at the zero lower bound but again it's not a death blow that there are things that the economy that policy makers can do, even monetary policy. Thank you, you're welcome. I'm also a master's student at law school. So here comes the second question from our audience. In 2008, we saw a large increase in oil price. So the question asking, does controlling for oil price shocks affect these results? We don't know, so we've never done that. You're absolutely right. So this is, I kind of was a little quick and dirty in case there were non-technical people in the audiences about what our regression procedure was. So the way I described it is we were using, we have lots of countries, lots of financial crises, what happens on average after these. And I described that in thinking about what happens on average we control for, what happened in output before, a country specific factor. We also include a time fixed effect. So basically a specific factor for each quarter, precisely to take account of kind of, things that are happening worldwide in each period. And so to the degree that oil prices are there, we would have controlled for that. I think, again, one of the things we didn't, so I think for what we found before oil prices are not gonna be a big part of the story, sort of thinking about as we go and think more about 2008, sort of the fact that oil prices went up and then went down, sort of thinking how that sort of factors in could be interesting. This question comes to us from Twitter. What explains the sudden increase in crises across the OECD in the 90s? It really jumps out at the wheel. In the 1970s and 80s and a lot of them in the 1990s. I mean, this is not something, again, that we have a particularly informed opinion about. I'm not saying we've done research on. I think I'm probably in the school of thought that the other thing that goes on is sort of financial deregulation and financial innovation. Sort of both were happening big time in the 1990s. And I think part of what exactly went wrong is innovation. So innovation not only did regulation not keep up with innovation, at the same time we're doing all this innovation, we're deregulating. And David often describes, we're reading these OECD documents and the whole thing is that they're written in real time. And it's really a little unnerving like you're reading about Sweden or you're reading about Norway and before their crises, it's we're doing great things. They're liberalizing their financial markets. They're getting rid of all these regulations and we as modern people knowing what happens, like no, don't, don't. But anyway, so I think that is a big part of sort of what went on and why the numbers look or why the crises did sort of have a resurgence. So this question has been, one factor in the current crisis is a prolonged effect in terms of unemployment rate. Although the GDP is recovering but the unemployment rate is still high. So can you tell us something more about the unemployment rate? Well, I guess as we tended to look mainly at industrial production and GDP in part because it's sort of seems, it's more consistent over countries and how you measure unemployment rates and things are somewhat more difficult and how to make that consistent across countries. But absolutely, one of the striking sort of facts about this recession has been just how long the unemployment rate has stayed high. Importantly, I guess having kind of said all that I think I tend to come down. So there is a tendency to say, as you sort of as the questioner kind of phrased it, well GDP growth has been perfectly fine but unemployment has stayed really high. And I think a key fact is relative to sort of where we were GDP growth hasn't been okay, right? So we have gone back to kind of normal GDP growth. So if you think normal for the US is two and a half percent a year, yes, we've been doing that. But normally when you've had this terrible recession you'd have a period where GDP would grow 4%, 6% coming out of the Great Depression, it grew 10% a year. And that exactly what we haven't had this time is that period of really robust GDP growth. We just kind of went back to at best normal and often below normal. So I think the main reason why unemployment has stayed so high is because we have not had this really aggressive sort of recovery of GDP. But then there are certainly other factors beyond that. So there are I think one of the things that has certainly worried me both as a policymaker and then as someone watching policy, just because this recession has dragged on so long part of what that does is then create a core of people that have been unemployed for a very long period of time. And one of the most distressing things is that it becomes harder and harder for them to ever get a job. And so this what we often is called in the literature these hysteresis effects that a period of very high unemployment may just make the normal unemployment rate higher. And I think that is certainly one of the things that has made me very concerned and I worry we are seeing some of that going on now. Due to the underlying composition of affected economies especially in heavily financial actually. That logic right which you'd wanna be saying. So like in this particular episode right so the United States we have a big financial sector so the financial crisis has a particularly big impact. I guess I mean I don't think I have a informed answer on that. My sense is you know, huh. David do you wanna answer this one? Darn. The United States and the UK in the current episode. But to your point like Sweden was right. You know a lot of the things like the Sweden night I did the example so. Goes down and I did this nice song and dance about they recognize their losses they recapitalize the banks. They had six banks right so it's really easy to kinda do that. I mean six big banks but still. You know it's not like trying to do the American financial system and so but that's a way of just doing concretely these countries having financial crisis a lot of them in the 90s were not your big financial centers. So anyway so I don't have a good sense of whether that might be explaining what's going on. It's not something we've looked at yet. So for the government or administration what steps can be taken to pack away for a better economic policy? I think the you know I have a very I often tell a story that for any academic in the audience the first time I sent a paper to a journal like within a month I got a response that said we love your paper we want to publish it. That never so I thought wow publishing is really fun right never happened again right never ever ever since then it's like if you're lucky you get a painful revise and resubmit. But the same is true of policymaking right so think about my time in Washington it's like I went there we passed the recovery act within a month and then we did health reform legislation and then we did you know financial regulatory reform so I had this vision like you know and you know I often tell this story I mean it was a fantastic time to make policy because you know one of the things the president was very much he's an evidence based kind of guy right so the way you want an argument with them is to have smart people like Josh working for you at the CEA and say get me the best evidence we have right and so we would just like do studies and run the CPS tapes and it resonated and you often won arguments that way and got to really influence policy. You know the trouble that's how policy ought to be made right it ought to be made on the basis of evidence and of people looking at the evidence dispassionately and saying just what's true and what's right and for a brief shiny window it did feel like it was like that. I think it's much less like that now and I don't know what the answer is other than to I mean this is normally where I'd say please go out and vote for evidence based policy making right that's I think that's the only you know the only way you can do that and the thing I find the most frustrating is you know people being feeling free to say you know things that just aren't true or where there are 10 studies on one side and at most one bad one on the other side and that doesn't matter that doesn't seem to carry the day and I don't know of any way to make it carry the day other than for voters to insist on it but that's I think the ultimate of. Okay so this is our last question. What advice do you have for women in male dominated fields like economics and business? Well since I have my husband sitting here I'll say marry well. So really I mean truthfully I mean I think one of the things that has you know made it possible for me to do life as an academic and to work in government is having a supportive spouse and I often this is a terrible story and please don't ever tell the IMF but David when I was going to Washington claims that he walked around Washington saying I'm looking for the dullest job possible because and he went to work for the IMF. So sorry you're in such trouble. But it was right we kind of knew we had a son at home I was gonna have a job that was 80 hours a week and the only way I could ever possibly do that is if he was willing to be the one that was always home meeting the plumber and doing the laundry and for two years that was what he did. And so I mean I think that you know it is a partnership one of the things I often claim is I really want him to go have to do the job and I want to be at home meeting the plumber because sometimes the policy job was not the most fun. But anyway but I think you know so I think that is certainly a big part of it. The other thing that really struck me and I don't quite know sort of how I feel about this Berkeley is and I would guess it's the same at Michigan Berkeley is a uniquely well adjusted place so being a woman just never kind of I didn't notice. It was just it was a happy collegial place and I don't think anyone else noticed that I was a woman. Washington was definitely different and I definitely found it stressful and it sometimes felt like it was a club and I was knocking on the door. And the one thing that the women in the White House so one of the nice things is there were a lot of women in the White House. We just started having dinner together every you know couple of weeks and it mattered just this sense of you know there was a little bit of complaining and there was a lot of mutual support and I think that was important but it also became our network and so some of you know sometimes when I was not in a meeting and I suddenly showed up on the manifest I knew that there was some someone watching out for me and more often than not it was one of the other women in the White House and I think that is probably you know it's a shame we have to be that way but I think that mutual support you know whether it's a group of women or just a mutual support of colleagues, of mentors those kind of things I think that matters a lot so that's the best advice I can give you. Anyway thank you so much this has been really wonderful for me. Well I'd like to thank Christy for her hot off the computer new work and all the implications that she shared with us. I'd also like to thank all of you for coming and all of the questions. I know we didn't have time for all of them so I hope you will stay and join us for a reception just in the back in the room here and we can continue the conversation. So again thank you for joining us for the policy talks at the Ford School.