Stability does not produce macro fragility

Here’s Matthew Klein over at Free Exchange:

I am sceptical that any central bank is capable of fulfilling its objectives over any meaningful length of time because, as the late Hyman Minsky explained, lower observed macroeconomic volatility in the short term encourages greater financial risk-taking. Thus, the longer the perceived good times last, the more fragile the economy becomes.

This is a popular, view, but I believe it has things exactly backwards.  Macro instability leads to banking crises.  Let’s review a few cases, and see how they were misinterpreted by pundits:

1.  Governor Strong’s policies produced macro stability from 1922-28.  A Great Depression started in August 1929.  Doesn’t that prove Minsky’s point?  Just the opposite.  The Fed adopted a tight money policy in 1928-29 to stop the stock market boom.  This policy cut NGDP in half by 1933, making a severe financial crisis inevitable.  Here’s the key fact that proves Minsky’s model doesn’t apply to the Depression; there was no financial instability during the severe slump of August 1929 to October 1930, a slump that was much deeper than the Great Recession.  If the financial system in 1929 had been “fragile” it would have collapsed like a house of cards in the deep contraction of 1930.  But it didn’t, indeed the first really severe banking crisis didn’t begin until mid-1931, and most big banks survived the 50% fall in NGDP–an incredible performance that no modern big bank could match.  The problem today is not macro stability; it’s FDIC and TBTF.

2.  There was a long period of stability from 1950 to 1973.  It did not make the macro economy unstable.  The problems of the 1970s were partly supply side–leading to modestly lower RGDP growth.  The main problem was high NGDP growth leading to high inflation.  None of this had anything to do with the financial system.

3.  The period from 1983-1990 was also fairly stable, and was followed by a banking crisis in the US (mostly S&Ls). But the recession of 1991 was pretty mild, nothing at all unusual.  And that recession can be fully explained by the Fed’s decision to reduce inflation from 5% to 2%; banking played no significant role. The 1991 recession was no worse than the preceding half dozen recessions, which saw no banking crises.  BTW, for those who (wrongly) think I’m a “dove,” I’d point out that the main cause of the 1991 recession was not money being too tight in 1991, but rather too loose in 1987-1990.  Too much Volcker stimulus (and now Volcker thinks Bernanke is too easy!)  This is one recession that the Austrians were basically right about.

Update:  Doug pointed out Greenspan took over in 1987—my mistake.

4.  The recent recession was of course caused by highly unstable NGDP growth. There might be a grain of truth in the Austrian claim that money was too easy during the housing boom–but only a grain.  At worst it was slightly too easy.  On the other hand the far too tight monetary policy of 2008 caused a huge drop in NGDP growth, to 9% below trend.  That’s what caused both the recession, and the intensification of what had until mid-2008 been a manageable financial crisis. Most people have reversed causality.

In 1929 the Fed departed from Gov. Strong’s macro stability policy and moved toward a financial stability target.  The results were disastrous. Let’s never make that mistake again.


Tags:

 
 
 

50 Responses to “Stability does not produce macro fragility”

  1. Gravatar of Luis Enrique Luis Enrique
    1. March 2013 at 07:18

    here in the UK we know how to resolve such questions

    http://www.youtube.com/watch?v=Np6gyUb0E7o

    there’s only one thing for it, Scott Sumner and Ashwin Parameswaran FIGHT!

  2. Gravatar of Geoff Geoff
    1. March 2013 at 07:21

    “…lower observed macroeconomic volatility in the short term encourages greater financial risk-taking. Thus, the longer the perceived good times last, the more fragile the economy becomes.”

    “This is a popular, view, but I believe it has things exactly backwards. Macro instability leads to banking crises.”

    Why can’t BOTH views be true?

    What if it’s the case that macro stability caused by non-market intervention AND macro instability caused by non-market intervention, both lead to banking crises, because banks require neither non-market caused stability, nor non-market caused instability, but market caused stability or instability?

    Wanting a force-based “stability” is a chimera in social life dating back to at least the middle ages. The idea of perfection was no change. Constancy. Perfect beings do not change, lest they change away from perfection. So that idea of perfection was manifested in social life on Earth, where no change was SOUGHT after.

    Except human action IS NOT STABLE. Humans learn. Humans change.

    There should be neither stability nor instability caused by anything other than preference and knowledge changes.

    MM suffers from the same plague of wanting “no change” as the middle age priests and mystics. For MM, the desired “no change” is aggregate spending growth. That’s the new God.

    I’m not convinced.

  3. Gravatar of Master of None Master of None
    1. March 2013 at 07:38

    And yet, could such large and deep recessions occur without “stable NDGP” central bank policies, or rather, any central bank policies?

    Posts like this always bring my mind to the question of whether or not we should have a central bank – something that Steve Randy Waldman was hinting at in his recent post: http://www.interfluidity.com/v2/4043.html

    Scott – your arguments for stable NGDP are pursuasive, but our experience over the last 100 years has proved that the current system can repeatedly fail to achieve its goals. Do you deny this?

    I would like you to address this question more often: Should we have a central bank? Does the answer to that question depend on our belief that the central bankers will actually maintain stable NGDP growth? Should we believe that they will?

    I cannot believe that a referendum on having a central bank would pass today, given our experience with them. And why should it?

    Wouldn’t more frequent, but less painful, recessions be fairer to each generation? It would allow every generation at least a couple of chances to succeed in a positive economic cycle. We only get one shot at life; it is terribly unfair to artificially impair every 6th or 8th generation’s lifetime earnings (regardless of good intentions).

  4. Gravatar of effem effem
    1. March 2013 at 07:41

    I couldn’t disagree more. As an actual market participant, I can tell you that any perceived decrease in economic volatility will be met with increased leverage. If my revenues/profits grow more smoothly I will lever them up to the overall volatility target I am comfortable with. I believe this is exactly why the financial sector massively increased leverage during the Greenspan era as the “Greenspan put” served to reduce economic volatility.

  5. Gravatar of OhMy OhMy
    1. March 2013 at 07:48

    “Macro instability leads to banking crises.”

    Minsky is not only about banks. Banks can survive, indebted private sector might not. In 2008 the households were indebted and reached the so called Ponzi stage, in 1928 it was firms.

    Unlike Minsky, you have no mechanism for a crisis. Only after the fact you can say: hey, there was a crisis because NGDP was unstable. Ahem, falling NGDP is the definition of a crisis. So you are saying: “there was a crisis because there was a crisis”. Great. Do you have any mechanism for a crisis in an economy without a central bank?

  6. Gravatar of Suvy Suvy
    1. March 2013 at 07:53

    “Stability does not produce macro fragility”

    If you define aggregate demand as income plus the net change in debt, everything that happened including the S&L crisis and other banking crises becomes easily explainable.

    “There was a long period of stability from 1950 to 1973.”

    This was basically during the post-war era after a 15-20 year depression that ended around 1943ish. Of course the next 20-30 years will be relatively stable.

    “The period from 1983-1990 was also fairly stable, and was followed by a banking crisis in the US (mostly S&Ls). But the recession of 1991 was pretty mild, nothing at all unusual.”

    This is because private sector credit demand was still relatively high. Just look at debt/income ratios across all sectors(especially the financial sector), the financial crises since 1980 have gotten progressively worse and we’ve had slower and slower recoveries as debt/income ratios have increased over time since the post war era. From 1950-1973, unemployment remained relatively low and there were constant fluctuations between unemployment and inflation. From 1980 on, there have been virtually no fluctuations between unemployment and inflation while debt/income ratios exploded.

    One very common thing about natural systems is that they usually benefit from some form of volatility. Nassim Taleb talks about this heavily in Antifragile. With natural/organic systems, they communicate with their environment via stressors. If you think about such a system and try to remove the natural volatility, you’re increasing the chance of a blow up because you’re not letting the system communicate with its environment.

    Trying to overstabilize an economy is like trying to micromanage forest fires. When you have micromanage forest fires, it allows deadwood to build up and creates worse forest fires later on. With an economy, you get a build-up of debt/income ratios which cannot sustainably grow forever. This leads to major blow-ups/depressions.

  7. Gravatar of Suvy Suvy
    1. March 2013 at 07:57

    “The recent recession was of course caused by highly unstable NGDP growth.”

    No, that wasn’t the cause. That was a consequence of bad policy, but it was not the cause. The cause was a buildup of debt/income ratios. If you make the (seemingly) small differing assumption that the creation/destruction of private sector credit is not demand neutral, you get very, very different results. Especially when you assume that this growth in credit doesn’t have to show up in consumer prices, but can show up in the price of an asset sector.

    If you literally make one differing assumption, you get very, very different results.

  8. Gravatar of Suvy Suvy
    1. March 2013 at 08:04

    Effem,

    “I couldn’t disagree more. As an actual market participant, I can tell you that any perceived decrease in economic volatility will be met with increased leverage. If my revenues/profits grow more smoothly I will lever them up to the overall volatility target I am comfortable with. I believe this is exactly why the financial sector massively increased leverage during the Greenspan era as the “Greenspan put” served to reduce economic volatility.”

    That’s Minsky’s exact point. If you go through the model that he uses, he actually has shifting yield curves of market participants as economic conditions change. The future expectations of volatility are the driving factors behind his yield curves.

    I think a good way of thinking about it is that leverage is a great thing when times are good and a horrible thing when times are bad. If the market as a whole expects time to be good, the entire market will lever up. The only thing that needs to happen is that the market’s expectations have to be wrong–which happens a good part of the time. Expectations aren’t driven by “rationality”, their driven by human psychology, which is unpredictable, has tendencies of herd behavior, groupthink, euphoria, depression, etc. Financial markets are just all of those aspects of human psychology aggregated all at once.

  9. Gravatar of Stability does not produce macro fragility | Fifth Estate Stability does not produce macro fragility | Fifth Estate
    1. March 2013 at 08:10

    […] See full story on themoneyillusion.com […]

  10. Gravatar of Jon Jon
    1. March 2013 at 08:31

    The Austrian critique applies to any level targetting regime. Level targetting should be a special purpose tool: Reserved for the most severe of past policy errors–those where the delta from the expected level path is too severe to heal with a couple quarters.

  11. Gravatar of Niklas Blanchard Niklas Blanchard
    1. March 2013 at 08:35

    I wrote a similar post here, at Miles Kimball’s request. But I think that TB2F is massively oversold, and the real story is “too brittle to sustain”, or TB2S, as I like to call it.

    I’m kind of surprised you would latch onto TB2F.

  12. Gravatar of Michael Michael
    1. March 2013 at 08:49

    “One very common thing about natural systems is that they usually benefit from some form of volatility. Nassim Taleb talks about this heavily in Antifragile. With natural/organic systems, they communicate with their environment via stressors. If you think about such a system and try to remove the natural volatility, you’re increasing the chance of a blow up because you’re not letting the system communicate with its environment.”

    Under an NGDP level target you would have volatility in real GDP growth and inflation.

  13. Gravatar of Chris Chris
    1. March 2013 at 08:54

    I know that in Britain esp, bank leverage (assets:equity) increased by something like 30% between 2000 to 2007. I’m not sure why, I once speculated it could have had something to do with Glass-Steagal being repealed in 99 but the link was tenuous at best.

    In general I’m quite skeptical of models that would attribute this behavior to over optimism and say that an unavoidable crisis will result from it. Once you have a better model than the one everyone else is using, they’ll start using you’re model, making it redundant.

    There are philosophical issues regarding whether we can predict far in advance a coming financial crisis because the knowledge of that prediction changes the very nature of what we’re observing.

  14. Gravatar of Suvy Suvy
    1. March 2013 at 09:00

    Michael,

    “Under an NGDP level target you would have volatility in real GDP growth and inflation.”

    I agree. I support a NGDP target of 4-5%.

  15. Gravatar of Max Max
    1. March 2013 at 09:33

    I find it strange that Market Monetarists never talk about GDP volatility (or volatility of volatility). A fanatic devotion to GDP targeting practically guarantees that there will be, from time to time, wild oscillations around the GDP target path. The central bank will miss big on one side, then the other, before things eventually settle down again. That is not economic stability even though the CB is doing its job 100% correctly.

  16. Gravatar of ssumner ssumner
    1. March 2013 at 09:47

    effem, You said;

    “As an actual market participant, I can tell you that any perceived decrease in economic volatility will be met with increased leverage. If my revenues/profits grow more smoothly I will lever them up to the overall volatility target I am comfortable with. I believe this is exactly why the financial sector massively increased leverage during the Greenspan era as the “Greenspan put” served to reduce economic volatility.”

    That’s true, but faulty government regulation also played a role.

    OhMy, You said;

    “Ahem, falling NGDP is the definition of a crisis.”

    Then I guess we don’t ever have to worry about Zimbabwe having any financial crises.

    Suvy, You said;

    “If you define aggregate demand as income plus the net change in debt,”

    Why would you do that?

    Jon, Level targeting can’t be “reserved,” as you’d no longer have level targeting in that case.

    Niklas, I’ve argued that FDIC is a much bigger problem than TBTF.

    Max, You said;

    “I find it strange that Market Monetarists never talk about GDP volatility”

    ?????

  17. Gravatar of Suvy Suvy
    1. March 2013 at 09:48

    By the way, I found a couple of good articles on a different view of a banking system. Banks are not constrained by reserves; they are constrained by the costs of lending.
    “They arise from the costs of lending, which is conditioned by (a) the interest rate targeted by the Fed, (b) regulatory and market capital requirements and the market price for bank capital, (c) by back-office administrative and hedging costs of lending, and (d) the credit-worthiness and credit-hungriness of borrowers.”

    So shifts in expectations about macroeconomic volatility have a massive role in the amount of credit(and thus money) in the system. Here’s the article:
    http://www.cnbc.com/id/46970418

  18. Gravatar of nickik nickik
    1. March 2013 at 09:58

    > lower observed macroeconomic volatility in the short term encourages greater financial risk-taking. Thus, the longer the perceived good times last, the more fragile the economy becomes.

    > I couldn’t disagree more. As an actual market participant, I can tell you that any perceived decrease in economic volatility will be met with increased leverage. If my revenues/profits grow more smoothly I will lever them up to the overall volatility target I am comfortable with.

    I actually agree with that stability will increase reserve ratios but thats not a bad thing per se. In the time of the gold standard some private companys sold 100 bonds.

    If we can control the austrian and the MM cycle (NGPLT should take care of that) we should only have supply shocks to worry about. Now private companys are well aware of supply shocks and worry about them a lot. A system that is stable for a long time has to deal with a lot of big and small supply shocks and learn to get used to them.

    Look at the Free Banking Systems, with time there reserve ratio become smaller, I belive Selgin has shown that the scottish banks only held 2% reserve ratio but the system was quite stable. These banks even had shareholders that where liable up to the double value of there stock (or something like that).

    There is no reason to think that NGDP will be hard to hit even when banks there is more risk taking going on. Especially a forward looking target, like NGDP future.

    One thing I would agree to is that the long the NGDP target holds, the bigger the upset will be when the target is not missed for a long time. Remember with a level target you can easly be long term stable without always be perfectly correct. The gold standard was not short term stable eiter, only long term and it worked fine.

    I doute that private risk taking can ever be so out of control that it is impossible for a bank to hit a avg. growth level over some years. After some time the real supply shocks become the more importend worry, the constraint is now supply shocks.

    The real problem, and why I do not want central bank, is political. Even if you have the perfect MM central bank, it can easly be taken over again. Also I belive a free banking system would perform just as well.

  19. Gravatar of Doug M Doug M
    1. March 2013 at 10:04

    “there was no financial instability during the severe slump of August 1929 to October 1930”

    Really? Really!

    “The period from 1983-1990 was also fairly stable”

    Except that 1987 was one of the most volitile year in the last half century.

    “…the main cause of the 1991 recession was not money being too tight in 1991, but rather too loose in 1987-1990. Too much Volcker stimulus…”

    Greenspan was appointed Fed Chairman 6/1987.

    “The recent recession was of course caused by highly unstable NGDP growth.

    NGDP did not decline until after the finacial markets cracked.

    Least volitile years….as measured by the standard deviation of returns of the S+P 500 1955 to present.

    1964, 1995, 2006, 1993, 1972

    Most volitile years:

    1987, 1974, 2009, 1998, 2008.

    Minsky doesn’t say that there cannot be extended periods of calm, but it is laying the groundwork for for a potentially volitile situation. It is not entirely unlike Hayek’s theory that the seeds of each recession are sown via malinvestment during the previous boom.

  20. Gravatar of Suvy Suvy
    1. March 2013 at 10:21

    Prof. Sumner,

    “Why would you do that?”

    Simple, credit creates purchasing power. If I swipe my credit card to buy a $1000 laptop; that $1000 doesn’t take away from someone else’s spending power. Finance is a revolving fund; it doesn’t depend on savings. It goes with the link that I posted above and the banks in macro argument as well. When banks issue credit, it creates money and adds to demand in the process. In reality, all forms of money are debt.

    As Keynes talked about, it’s a myth to think that ex-ante savings creates investment. The “finance” for the planned activity has to take place before the actual investment phase which then results in the incomes which are then saved;. Finance is a revolving fund and investment is not determined by savings. Savings is determined by the level of investment. Ergo, if aggregate demand is consumption plus investment(assuming net exports=0), then aggregate demand is income plus the net change in debt.

  21. Gravatar of 123 123
    1. March 2013 at 10:32

    @Max
    Yes, market monetarists are talking about vol:
    http://marketmonetarist.com/2012/12/22/guest-post-market-montarism-and-financial-crisis-by-vaidas-urba/

  22. Gravatar of Kailer Kailer
    1. March 2013 at 10:34

    In the long history of coin flipping we have seen it play out again and again. Long streaks of heads are inevitably followed by tails. Recall the time from the 190th to 200th coin flips where we saw an unprecedented 10 consecutive heads. Such good fortune can not be accrued without consequences, and sure enough, three of the following four coin flips were tails the most tails in a four flip span in over 25 flips.

  23. Gravatar of Geoff Geoff
    1. March 2013 at 10:39

    Michael:

    “Under an NGDP level target you would have volatility in real GDP growth and inflation.”

    Under being strapped to a tree, there would be lots and lots of eyelid volatility, lung volatility, and a whole bunch of other bodily volatilities.

    Clearly Taleb would agree with you that someone tied to a tree would have an increasingly good life due to all the volatility there.

    /s

    Moral of trite story: Volatility in country level NGDP is associated with a healthy society, the same way volatility in all sub-country level nominal spending categories are associated with a healthy society.

  24. Gravatar of Becky Hargrove Becky Hargrove
    1. March 2013 at 10:42

    Moving towards financial stability targets always implies a lack of belief in incremental growth potential, on the part of entire populations. Financial stability targets are also pure laziness on the part of those who don’t want to make loans unless they are for the largest potential wealth structures possible. But then any time wealth can’t realistically be added on in such a manner, finger pointing ensues to blame the ones who lost out at the high stakes table. And yet, the high stakes table is practically the only table many local economies either want, or offer to their participants.

  25. Gravatar of errorr errorr
    1. March 2013 at 10:44

    I don’t see that as an argument against what Minsky was saying. I see it as an argument that when the Minsky moment occurs a stable NGDP growth path will limit the losses and prevent a recession from being severe. Although people will freak out when the inflation hits even though it could never get bad enough to be unsustainable.

  26. Gravatar of Bill Woolsey Bill Woolsey
    1. March 2013 at 10:52

    So, nominal GDP grows smoothly for a long period of time.

    “Market Participants” “lever up.”

    So what?

    Things “go worse” than expected. They become insolvent and default. Their creditors receive partial payment.

    How does this prevent nominal GDP from continuing to grow as usual?

  27. Gravatar of Michael Michael
    1. March 2013 at 10:52

    “If I swipe my credit card to buy a $1000 laptop; that $1000 doesn’t take away from someone else’s spending power.”

    No, but it adds to the income of the seller.

  28. Gravatar of 123 123
    1. March 2013 at 11:58

    @Bill Woolsey

    When defaults start to increase, VIX explodes. By arbitrage, NGDP expectations of economic actors become more uncertain. The probability that NGDP will temporarily overshoot or undershoot the target goes up.

  29. Gravatar of Tyler Joyner Tyler Joyner
    1. March 2013 at 12:06

    Bill Woolsey,

    Something that I’ve been thinking about is whether NGDPLT would create incentives to over-lever. If NGDP is growing at a rate higher than the target, then inflation will be reduced to “target the forecast”. Reduced inflation, or outright deflation, increases real returns on bonds. Demand for bonds increases as returns go up.

    Am I missing something?

  30. Gravatar of nickik nickik
    1. March 2013 at 12:13

    @Geoff

    The way you look at the world is that every variable need to be influx always?

    How about a free banking system (based on a stable reserve currancy)? It would by all we know creat a more of less stable nominal spending as well but nobody ‘targeted’ it.

    Now is that good or bad?

  31. Gravatar of Bill Woolsey Bill Woolsey
    1. March 2013 at 12:18

    “over lever” means issuing and sellng bonds. It doesn’t mean buying bonds.

    If you think nominal GDP is currently “too high,” then buying bonds might well be a good idea. An alternative to buying bonds would be buying capital or consumer goods. Buying bonds rather than capital goods or consumer goods does tend to lower real interest rates, but it also directly results in less current spending on output. It dampens the increase in nominal GDP.

  32. Gravatar of ChargerCarl ChargerCarl
    1. March 2013 at 12:35

    Scott, in retrospect do you think cutting the inflation rate from 5-2% was a wise move? Obviously our preferred MP regime is NGDPLT, but a higher inflation rate would have meant higher trend NGDP growth and, consequently, higher equilibrium interest rates. This would have given us more room for rate cuts in 2008.

    Isn’t this how Australia avoided recession?

  33. Gravatar of Suvy Suvy
    1. March 2013 at 12:36

    Michael,

    If I use the debt to purchase a good or service, then yes. So rising debts cause rising incomes, but if debts rise vs incomes, that rise is unsustainable. This becomes a huge problem when the debt is used to finance large purchases of assets, like houses. Not only do I define effective demand as income plus the change in debt(where income and the change in debt are flows), but the money can be spent on goods and services or financial assets. This model gives you the longer wave debt cycles and “deleveragings” that you see in economic time series. As time goes on, debt/income ratios rise and they end up having to correct. We’re in one of those correcting phases as debts cannot sustainably rise relative to incomes forever. Not only that, but this model also allows for changes in the money supply to show up as changes in asset prices and not just price inflation. This model treats the asset sector separately from the real sector(goods and services).

  34. Gravatar of TravisV TravisV
    1. March 2013 at 13:06

    I just found this recent essay by Prof. Sumner:

    “The Neoliberal Revolution”

    http://reason.com/archives/2013/01/22/the-neoliberal-revolution

    Awesome read!

  35. Gravatar of marcus nunes marcus nunes
    1. March 2013 at 14:15

    “The recent recession was of course caused by highly unstable NGDP growth. There might be a grain of truth in the Austrian claim that money was too easy during the housing boom-but only a grain. At worst it was slightly too easy. On the other hand the far too tight monetary policy of 2008 caused a huge drop in NGDP growth, to 9% below trend. That’s what caused both the recession, and the intensification of what had until mid-2008 been a manageable financial crisis. Most people have reversed causality.”

    No, mon. policy was not even “slightly too easy”. It had been overly tight previously and in that period greenspan successfuly ‘made up’:

    http://thefaintofheart.wordpress.com/2012/04/02/was-monetary-policy-easy-in-2003-2004-not-according-to-robert-hetzel/

  36. Gravatar of ssumner ssumner
    1. March 2013 at 14:44

    Doug, I corrected the Greenspan mistake, all your other points are wrong. Don’t confuse financial crisis with stock market instability. But if you insist on doing so, let me point out that 1988 and 1989 were boom years, so financial turmoil obviously doesn’t affect the macroeconomy. Which is my point.

    I do agree that periods of instability often follow periods of stability. How could it be otherwise?

    Suvy, Extra purchasing power already shows up in income–you are double counting.

    Kailer, Coin flips are independent, if you are talking about the business cycle.

    errorr, I was commenting on Free Exchange, not Minsky.

    Charger Carl, Yes, it was a wise move, although in retrospect a higher inflation rate would have been better. Wise, given the reasonable assumption that future Fed policy would be sensible.

    Travis, Did I forget to link to that?

    Marcus, I still say it’s a judgment call, as no two economists agree on exactly where to draw the trend line. And recall that inflation was over 2% during the boom period–it should be under 2% during booms.

  37. Gravatar of Suvy Suvy
    1. March 2013 at 15:25

    “Extra purchasing power already shows up in income-you are double counting.”

    Not if you treat it as a dynamic system where income is a flow. It’s not double counting. You have to think in terms of ex-ante and ex-post. Ex-ante, it is debt, ex-post, it is someone else’s income. When someone else gets that income, they can borrow on that income stream as well. This is how money gets created and how incomes result. Everyone’s incomes come about the this way.

    There’s actually a way you can mathematically prove this based on a set of axioms assumed in the model. It requires a theorem from Lebesgue integration, which is an alternative to the traditional Riemann integration. It has to do with the way the limits approach from the left and right. The limits from the left side and the right sides of the spending(demand) curve aren’t the same. This is why you have to resort to Lebesgue integration in order to prove the theorem.

    The key point is that debt is added at discrete points, which then turns into someone else’s income. For example, if I swipe a credit card, that’s a discrete injection, not a continuous injection of cash. That point is absolutely critical and cannot be overlooked.

  38. Gravatar of Suvy Suvy
    1. March 2013 at 15:31

    The way we calculate spending is through effective integrating(or summing up) the amount of transactions that take place. You’re not double counting primarily because the injections of debt are discrete. They are injected at particular points in time and you are injecting money that wasn’t there before.

  39. Gravatar of Suvy Suvy
    1. March 2013 at 15:36

    The critical point is because the debts are injected discretely, incomes are higher at the next stage than they were at the previous stage. The effective demand is the total amount of spending. So if you view income as a flow, a discrete injection of debt is money that is spent that wasn’t there before. Then, it turns into someone else’s income.

    Keynes addresses the same thing when it comes to investment and savings. There’s a difference between ex-ante and ex-post, which plays a critical role. The “finance” for the investment is injected discretely into the economy.

  40. Gravatar of Suvy Suvy
    1. March 2013 at 15:39

    As for the integration comment earlier, I should clarify why that matters. When you measure spending, you’re integrating over that period. If you’re integrating over a finite number of discontinuities, the limits from both sides of the integrating function will be the same. That’s why it would seem like double counting, but its actually not.

  41. Gravatar of Doug M Doug M
    1. March 2013 at 22:02

    Instability was your word and I called you out on that.
    as in:

    “There was no financial instability during the severe slump of August 1929 to October 1930, a slump that was much deeper than the Great Recession.”

    Black Tuesday wasn’t financial instability?

    If you have a different defention of instability, then you better tell me what it is! Do you solely mean bankruptcy or bank failures when you say “financial instability”?

  42. Gravatar of Stability does not produce macro fragility « Economics Info Stability does not produce macro fragility « Economics Info
    2. March 2013 at 05:00

    […] Source […]

  43. Gravatar of Suvy Suvy
    2. March 2013 at 10:10

    “But if you insist on doing so, let me point out that 1988 and 1989 were boom years, so financial turmoil obviously doesn’t affect the macroeconomy. Which is my point.”

    Wait, so you’re saying that private debt crises, government debt crises, inflation crises, currency crashes, et al don’t weigh on growth? Rogoff and Reinhart actually looked at over 800 years of data and they come up with completely different results. They specifically compare real GDP growth after a banking crisis vs real GDP growth after a regular downturn; they find something very, very different. Rogoff and Reinhart do the same thing for employment; not surprisingly, they find the same thing. Same thing goes for the governments balance of payments; it becomes much worse after financial crises(whether public or private) relative to a regular recession. I highly recommend their book; I think it’s very, very well written and entirely empirically based.

    To say that the size of debts don’t matter; that debts can keep being run up relative to incomes; that financial crises have no impact on the real economy(whether public or private) goes against what we actually see. Think about it, if the US government just kept running 10% fiscal deficits for the next 20 years and the US government’s bond yields start to spike while the entire government restructures its debts, are you saying that won’t have an impact on the real economy?

  44. Gravatar of Geoff Geoff
    2. March 2013 at 12:11

    nickik:

    “The way you look at the world is that every variable need to be influx always?”

    I was very clear that this is not what I am saying.

    Beneficial volatility is volatility contingent upon the changes to all individual’s voluntary preferences and knowledge in a context of respect for private property, i.e. the free market.

    The ground is individuals, not the concept itself. My view is that we should not strive to attain putting a concept into fruition, such as “stability”, or “volatility”, or whatever, but to strive for each individual to be free in making their own economic decisions, and to let the aggregate chips fall where they may, be they “stable” or “volatile.”

    I don’t want to live in a world that is volatile, or do I want to live in a world that is stable. I want to live in a world where I can do whatever I want, in a social context of everyone else doing whatever they want as well, which of course includes everyone not having to receive any unwanted object or physical activity from anyone else. This is the ideal, which of course almost certainly won’t be achieved in full, but can be increased if only people like you and everyone on this blog sought after it, and you are all capable of this.

    “How about a free banking system (based on a stable reserve currancy)? It would by all we know creat a more of less stable nominal spending as well but nobody ‘targeted’ it.”

    Define “stable currency.” Is it stable by spontaneous order considerations, or is it stable by coercive government interference considerations?

    “Now is that good or bad?”

    Define good and bad. I am not being coy here. I am being serious. Good and bad for me may not be good and bad for you. But there is a good and bad for us both if only we raised those issues.

  45. Gravatar of ssumner ssumner
    2. March 2013 at 15:21

    Doug, Most people define financial crises as debt crises. I have no objection if you want to include stock market crashes, as the evidence is overwhelming that stocks have only a tiny impact on AD, unless they reflect expectations of a change in NGDP growth, which the Fed could offset.

    Suvy, Financial crises would often be associated with deep recessions even if the causality went from economic downturn to financial crisis. Much of their data is from the gold standard period, when there would be an interest rate mechanism that’s missing under a fiat regime. Under the gold standard financial distress led to lower nominal rates on riskfree assets, hoarding of cash, and falling NGDP. Not necessarily true under fiat money where the central bank can offset that effect.

    I don’t really understand your debt–income—AD argument.

  46. Gravatar of Suvy Suvy
    3. March 2013 at 10:48

    Prof. Sumner,

    Reinhart and Rogoff go through all sorts of data. They look at post-war data as well along with other fiat economies that existed beforehand.

    One important thing to understand is the history of debt and capitalism. Historically, people started to use credit and debt before they started to use money(money originated from credit). Even during the days of the gold standard and much earlier, commerce used to be conducted by “letters of credit” and “bills of exchange”. All credit/debt is simply a promise to pay. The most important and critical factor to capitalism is the relationship between creditors and debtors, in other words, capitalism is an economy that is driven by debt. Look everywhere across the world, it is the places without a solid banking system/financial sector that are truly the poorest. This is because not having a strong financial system means that it is very difficult for commerce to take place. This is why every single model of a capitalist economy should have banks and debt–those are the two central parts of capitalism. In fact, that is, I think, the reason why high inflation and heavy deflation are so destructive–they destroy the relationship between debtors and creditors, and thus, the foundation of capitalism. Almost all commerce, even today, is transacted on faith and promises. If those promises cannot be kept, you will have a crisis that will impact the entire economy.

    As for effective demand equaling income plus the change in debt, the model is one where all are flows across time. The injections of debt are discrete, which do go back into the flow and do result in further incomes through the Keynesian multiplier. However, it is not double counting because ex-ante, I have a certain amount of income, which is profits plus wages. Demand is the total amount of expenditure, so the total amount of expenditure if I’m a worker is my wage plus the amount I can borrow from the bank to buy a good minus the amount I save. Yes, this money does go back through into the flow of an economy, but that effect occurs over time(it takes time for a sale to result into a wage and doesn’t happen immediately) through the Keynesian multiplier effect. However, as others get those incomes, it allows them to borrow more to conduct their own transactions(whether to buy consumption goods or investment goods).

    As for the Lebesgue calculus stuff I was saying up there, it has to do with the fact that increases in debt are discrete injections which means that AD is a discontinuous function with a finite number of discontinuities(it’s impossible for an infinite number of transactions to take place). This means that when you aggregate the amount of total spending that takes place at any given point is the amount of your income that you spend plus the amount of debt that you take out at that particular point(which is discretely done). When you integrate over a period to get the aggregate level of spending, you have to use the Lebesgue integral to show that it is integrable. Not only that, but because amount of debt is discretely injected, the limits from the right and left of the function will not be equal. This is why it looks like double counted, but really, it has not been counted before. Mathematically, it has to do with the way that the debt is injected.

  47. Gravatar of Geoff Geoff
    3. March 2013 at 11:10

    Suvy:

    “One important thing to understand is the history of debt and capitalism. Historically, people started to use credit and debt before they started to use money(money originated from credit).”

    This is a theory induced from a set of observations that David Graeber has made. It isn’t fact.

    A lack of identifiable observations of money preceding credit/debt does not mean that money did not precede credit/debt, the same way that a lack of identifiable instances of lifeforms that were once composed of some of the same atoms in your body, does not mean that there were no preceding lifeforms that were composed of some of the same atoms in your body.

    It would be wrong to say “Because I cannot identify any preceding lifeform that was composed of some of the same atoms as in my body, it means there were no such preceding lifeforms. I am therefore the first lifeform that is composed of these atoms.”

    What Graeber believes might be true.

  48. Gravatar of Suvy Suvy
    3. March 2013 at 11:26

    Geoff,

    That’s a perfectly fair argument. From all of the knowledge of history that I have read and that has been collected so far that I know about points to the idea that credit came before money.

    Would you say that’s a fair way to put it?

  49. Gravatar of ssumner ssumner
    3. March 2013 at 16:23

    Suvy, I’ve addressed the Rogoff and Reinhart claims in many other posts. Their data has little bearing on what happened to the US after 2008.

  50. Gravatar of Geoff Geoff
    3. March 2013 at 21:17

    Suvy:

    Yes.

Leave a Reply