Maynard s Revenge: Keynesianism and the Crisis Lance Taylor New School for Social Research
Maynard s Macroeconomics I Fundamental uncertainty Prices of assets vs. prices of goods and services Output = Income so that Investment = Saving Output Determination by effective demand, not aggregate supply, i. e. savings adjusts to equal investment via changes in income ( Say s Law not valid)
Maynard s Macroeconomics II Richard Goodwin: The level of demand is influenced by the income distribution. In fact there is a fairly clear distribution vs. demand cycle for the US economy but it became much less favorable for labor after 1980.
Maynard s Macroeconomics III Wynne Godley: Income = expenditure economy-wide but any sector can have income spending so that it borrows from or lends to the rest of the economy. There are interesting trends and cycles in sectoral net lending or borrowing. After 1980 household net borrowing and net lending to the US from the rest of the world trended strongly upward.
Keynesian Ontology Fundamental uncertainty (Not probabilistic risk you cannot put a complete probability distribution on future events, in part because you cannot conceive of them). Ecological examples: Easter Island, CFC and ozone hole Economics: Bernanke s great moderation circa 2006. He did not (could not?) have had any inkling of the crisis to come.
Financial Consequences I Under fundamental uncertainty financial actors have to base decisions on conventions which may be stable for a time, but then can change rapidly. Maynard s metaphor: the beauty contest as a device for social magnification. Wittgenstein s idea that knowledge is based on social conventions follows similar lines. The difference is that Keynes s economic expectations can change rapidly, e.g. the crisis.
Financial Consequences II In the housing price bubble, expectations and valuation models got built around a convention of steadily rising prices. When they stopped rising the calculations fell apart. Equity and housing bubbles had diverse causes, but deregulation and emergence of finance theory were surely important.
Prices of assets vs. prices of goods and services I A capitalist economy has two sets of prices: for assets and for goods and services. They do not necessarily move together. Inflation is always defined as the growth rate of an index of prices of goods and services. Inflation targeting is the principal obsession of central bankers one reason why interest rates were held low after the mid-1990s while asset prices went into bubbles.
Prices of assets vs. prices of goods and services II Asset prices movements can be highly irregular see next slides The equity price bubble breaks in late 1990s, housing bubble in mid-2000s The bubbles were driven by economic actors increasing debt to buy assets with rising prices they built up leverage to harvest capital gains.
Indexes of nominal and real S&P 500 and GDP deflator rescaled (2000Q1=100) 140 120 100 80 60 40 20 0 1950-I 1951-IV 1953-III 1955-II 1957-I 1958-IV 1960-III 1962-II 1964-I 1965-IV 1967-III 1969-II 1971-I 1972-IV 1974-III 1976-II 1978-I 1979-IV 1981-III 1983-II 1985-I 1986-IV 1988-III 1990-II 1992-I 1993-IV 1995-III 1997-II 1999-I 2000-IV 2002-III 2004-II 2006-I 2007-IV S&P 500 Real S&P500 GDP Deflator
Housing price indexes and GDP deflator (2000Q1 = 100) 250 200 150 100 50 0 FoF Index Shiller 10 Index GDP Deflator
Liquidity and leverage I For Keynes, liquidity meant money a liability of banks and an asset of its holders. By social convention it is always accepted as a store of value and a medium for transactions. Economic actors fall back on liquidity in the form of safe assets because of fear. But for traders, liquidity became debt (a liability) incurred to buy assets with rising prices a very old story. Not discussed by Keynes but by the Keynesian economists Hyman Minsky and Charles Kindleberger
Liquidity and leverage II Simplest balance sheet: P A A = D + E Leverage: L = P A A/E The return to an asset is the growth rate of its price, i.e. investors desire asset price inflation. Fluctuating growth of asset prices was built into nearly all the valuation models from finance theory used by traders.
Liquidity and leverage III But with stable debt, growth of P A means that leverage goes down an incentive to run up more debt to buy more assets. A decrease in P A makes leverage go up pretty soon creditors close in. To try to restore their positions traders started selling assets into a falling market and so drove prices down further deleveraging.
Liquidity and leverage IV The convention that asset prices will keep rising fell apart. There was a flight to liquidity in the form of safe assets, e.g. money instead of debt, in a Keynesian fallacy of composition. The crisis was exacerbated because mortgages had been securitized into derivatives by banks and sold on to asset holders.
Liquidity and leverage V The derivatives were valued by using models that became useless. Derivative prices collapsed when housing prices fell. Consequent deleveraging was the financial side of the crisis. But there are financial crises all the time. Why did the latest one propagate to the real side of the economy?
Effective demand Keynes s postulate: X = Output = Y = Income This identity is built into GDP accounting, as invented by Keynes around 1940 Simplest example: X = I + C and Y = S + C so that I = S. C = C(Y) or S = S(Y) so that I determines Y and X the principle of effective demand. In more complete accounting, exports and government purchases are injections of demand, imports and taxes are leakages
Distribution and demand I Consumption C is also likely to depend on income distribution. Easiest to work in macro with wage and profit shares of income. How are they determined? Begin with labor productivity = Output/employment Productivity rises when output ( capacity utilization ) swings up at the end of recessions (shaded).
0.015 Deviation of actual labor productivity from its trend (log of quotient) for the US business sector. 0.01 0.005 0-0.005-0.01-0.015 Recessions Log Dev from Trend
Distribution and demand II Wage share = Real wage/productivity. There are productivity and wage share vs. output cycles. Wage share falls coming out of recession, then rises toward the peak. The falling share leads rising capacity utilization (can be rationalized in a simple predator-prey model from mathematical ecology).
Distribution and demand III This distribution/demand cycle is an idea from Marx and the Keynesian economist Richard Goodwin However, the pattern changed after 1980 with much bigger falls and weaker recoveries in the wage share during the cyclical output upswing. This shift reflected a loss in labor s bargaining power and was part of the run up to the crisis.
Time series for capacity utilization and labor share 1.1 1.05 1 0.95 0.9 Recessions Capacity Utilization Labor Share (1992=100)
Distribution and demand IV The other side of the falling labor share coin was a rising profit rate after 1980 (next slide). Also note the downward trend in the real interest rate the Greenspan put justified by slow inflation. This helped support the equity and housing price bubbles noted above (an idea dear to Keynes s friend and rival Friedrich von Hayek) But financial deregulation was a more fundamental cause! It was pushed hard by Greenspan and Obama s current economic advisors.
Profit rate and real short- and medium-term interest rates 0.1 0.08 0.06 0.04 0.02 0-0.02-0.04-0.06 Profit Rate (After Interest and Taxes) Real Funds Rate Real 5 yr T-Bond
Net borrowing I For any individual or collective economic actor, Net borrowing = expenditure income = investment saving = injections leakages. The sum of net borrowing flows across all actors must be zero so that investment = saving economy-wide as in Keynes s national accounts.
Net borrowing II Look at the data for US households, business, government, and net borrowing from the rest of the world (a negative value signifies net lending to the USA): Household net borrowing was generally negative until around 1980, financing positive borrowing by government and business. Then it trended strongly upward consumption rose by around 10% of total income (the biggest component was spending on health care)
Net borrowing III The counterpart was falling net borrowing by the rest of the world, i.e. higher lending to the US to finance a rising external deficit. To a degree, this pattern reverses after the crisis. Also note that household borrowing tends to lead the business cycle, mostly due to rising investment in housing as the economy comes out of recession. Government net borrowing is countercyclical, most notably since 2007
Sectoral net borrowing flows normalized by GDP and NBER reference cycles. 0.12 0.12 0.07 0.07 0.02 0.02-0.03-0.03-0.08-0.08 Recessions Household Recessions Business 0.12 0.12 0.07 0.07 0.02 0.02-0.03-0.03-0.08-0.08 Recessions Government Recessions Foreign
Superimposed net borrowing flows normalized by GDP and NBER reference cycles. 0.12 0.07 0.02-0.03-0.08 Recessions Government Foreign Household Business
Lead in to the crisis I Political economy historical background: 1918-1929: rising income inequality and liberal (European sense of the word) ideology, leading into the Great Crash based on a run-up in leverage during the stock market boom. 1929-1950: Great Depression, New Deal (heavy financial regulation!), WWII, and recovery. 1950-1970: Creation of welfare state, Keynesian consensus, Golden Age of economic performance worldwide.
Lead in to the crisis II 1970-1980: Stagflation, oil shocks, loss of post WWII US international hegemony. 1980-2007: New liberal resurgence, financial deregulation, emergence of finance theory Rising inequality (top 1% of households have 22.5% of income in 1929, 9% in 1979, 22.5% again in 2006)
Lead in to the crisis III Next slide: along with income concentration there is the trend decrease in wage share noted above (though Goodwin cycle persists) A big increase in the consumption share of income A big increase in household real debt. In effect, most households displaced stable or falling real incomes into more debt to increase consumption!
Wage share of value-added, consumption share of disposable income, and household debt to income ratio with NBER reference recessions 110 200 107.5 175 105 150 102.5 125 100 100 97.5 75 95 50 92.5 25 90 0 Recession Consumption Share (left axis) Wage Share (left axis) Real Debt (right axis)
Lead in to the crisis IV The collateral for the debt was rising equity and (especially) housing prices. Household borrowing was eased by falling real interest rates. Note how it overshot fall in housing prices (typical pattern in financial cycles). However, the ratio of debt to net worth was fairly stable, and ratio of expenditure to net worth actually fell on trend until the crisis. One could say that consumers were borrowing rationally (NINJA mortgages and all) believing that housing prices would just keep on rising.
175 Real medium-term interest rate, housing prices, and real household debt with NBER reference recessions 0.1 150 0.08 125 0.06 100 0.04 75 0.02 50 0 25-0.02 0-0.04 Recession Real Housing Price (left axis) Real Debt (left axis) Real 5yr T-Bond (right axis)
Real US housing price index vs. real consumer debt 180 Real Housing Prices (2000-I=100) 160 140 120 100 80 60 40 20 1985-I 1980-I 1987-I 2000-I 1990-I 1994-I 1996-I 1998-I 1999-I 2003-I 2001-I 2002-I 2004-I 2005-I 2006-I 2007-I 2009-I 2008-I 0 35 55 75 95 115 135 155 175 Real Debt
Household debt and expenditure relative to net worth 0.3 0.24 0.25 0.22 0.2 0.2 0.18 0.15 0.16 0.1 0.14 0.05 0.12 0 0.1 195401 195504 195703 195902 196101 196204 196403 196602 196801 196904 197103 197302 197501 197604 197803 198002 198201 198304 198503 198702 198901 199004 199203 199402 199601 199704 199903 200102 200301 200404 200603 200802 Recession Household Debt / Net Worth (Left Axis) Household Expenditure / Net Worth (Right Axis)
Lead in to the crisis V Finally, after WWII the US had strong external surplus on trade and rising surplus on factor income The trade surplus disappeared around 1980, with Japan, Germany, China, and OPEC running surpluses instead. Global economic flows became highly unbalanced -- the external complement to internal imbalances
Lead in to the crisis VI Data on flows in the balance of payments are available post-1980. They are scaled to world GDP in the next slide. Look at downward trend in net borrowing, especially after early 1990s when the US exported military services for the Gulf War. It is offset by net short-term capital inflows (flows in minus flows out). US maintains net (though declining) long-term foreign investment and gets interest and dividend income from abroad.
US net international flows as % of world GDP (1980-2008) 2.50% 2.00% 1.50% 1.00% 0.50% 0.00% -0.50% -1.00% -1.50% -2.00% 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Goods and Services Interest and Dividends Unilateral Transfers Long-Term Investment Short-Term Investment and Reserves
Lead in to the crisis VII Bilateral current account deficits with Japan, Germany, China, and OPEC accounted for much of the US overall deficit. Japan was important in the 1980s, China and OPEC in the 2000s. China also emerged as the major provider of short-term capital after the mid-2000s).
US total and bilateral current account balances as % of world GDP (1980-2008) 0.20% 0.00% -0.20% -0.40% -0.60% -0.80% -1.00% -1.20% -1.40% -1.60% Total China Japan Germany OPEC
US total and bilateral balances of short-term capital and reserves as % of world GDP (including errors and omissions, 1980-2008) 2.50% 2.00% 1.50% 1.00% 0.50% 0.00% -0.50% China Japan Germany OPEC Total 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
Complexity of the crisis I So we get nine interacting sets of events: 1. Major shift in political economy after 1980 2. Real interest rate and wage share trend down after 1980 across cycles; profit rate and household borrowing trend up. Such trends cannot continue forever 3. Ratio of household debt to income doubles.
Complexity of the crisis II 4. Equity price boom from early 1980s to late 1990s, thereafter negative real equity returns. Housing price boom peaks in mid-2000s 5. Asset price bubbles allowed households to maintain consumption even with rising inequality. 6. The twin to rising household borrowing comprised foreign trade surpluses. Ben Bernanke argued that they resulted from a global savings glut in a well-functioning market; others see policy accommodation between the US and China in support of an overvalued renminbi.
Complexity of the crisis III 7. Falling real interest rates supported bubbles. 8. Light touch regulation was probably more important as a cause (US vs. Canada). It was rationalized by the emergence of high tech finance theory. 9. Keynesian macroeconomics was supplanted by Say s Law in the eyes of the economics profession; finance theory flourished.
Complexity of the crisis IV Factors 7 through 9 supported the financial mania, panic, and crash nothing new in historical terms. Shifts in household behavior (points 3 and 5) aided by asset price bubbles (point 4) were main transmission channel to the real economy. With the financial crisis, aggregate demand collapsed along with household net borrowing. Major recession Say s Law was not enforced
Complexity of the crisis V Global trade and financial flows adjusted to be consistent with crisis in the US The changing political economy environment made the whole process possible. ********************* Even Maynard Keynes could not have foreseen the whole chain of events, but the only way to understand it is in his intellectual terms that is the source of his revenge.
If you want to see more. Lance Taylor, Maynard s Revenge: The Collapse of Free-Market Macroeconomics, Harvard University Press, 2010