Skidelsky Warwick Lecture
In my third and fourth lectures examine the monetary and fiscal confusion which as reigned in the last five years -the experiments with ‘unorthodox monetary policy’ and the austerity drive in fiscal policy -as policy makers sought a path to recovery.In my fifth lecture 1 kick at the causes Of the crisis from the standpoint of the world monetary system.Finally, I ask the question: what should post-crash economics be like? What guidance should economics offer the policy-maker to prevent further calamities of the kind we have just experienced?
What should students of economics be taught? In this lecture I will consider only those bits of pre-crash orthodoxy relevant to policy making, tit main emphasis being on UK developments.
Theories of expectation formation played an overwhelming parting shaping the theory of macroeconomic policy; with changes in the way economists modeled expectations marking the different phases of theory. I will treat these in roughly chronological order, starting with the Keynesian theory. II.
UNCERTAIN EXPECTATIONS Keynesian macro theory dominated policy from roughly 1945-1975. The minimum doctrine -not in Keynes, but in accepted versions of Keynesian theory -to justify policy intervention to stabilize economies is: SLIDE 1 1. Uncertain expectations, particularly important for investment, leaving investment to depend on ‘conventions’ and ‘animal spirits’. 2. Relative interest-inelastic of investment. 3. A) sticky nominal wages (unexplained) and b) sticky nominal interest rates (explained by liquidity preference).
The first point suggested investment was subject to severe fluctuations; the last suggested there was a lack or weakness of spontaneous recovery mechanisms- ii the possibility of ;under-employment equilibrium’. This led to a prescription for macro-policy: to prevent or minimize fluctuations of investment demand. Point 2 in combination with b suggested primacy of fiscal over monetary policy for stabilization. SLIDE 2 ‘For Keynes, it was the tendency for the private sector, from time to time, to want to stop spending and to accumulate financial assets instead that lay behind the problems of slumps and unemployment.
It could be checked by deficit spending’. (C J. Also and D. Makes (1985), in D. Morris (De. ) ‘The Economic System in the UK”, 374) ‘In the standard Keynesian economic model, when the economy is at less than full capacity, output is determined by demand; and the management of economic activity and hence employment is effected by managing demand’. (ibid, 370) Mention in passing, that there was a theoretical and social radicalism in Keynes obliterated in the standard postwar Keynesian model.
For example, he thought insufficient demand was chronic and would get worse; and that, in consequence, the longer term survival of a free enterprise system depended on the redistribution of wealth and income and the reduction in hours of work. I will return to these points in my last lecture. Demand- management The government used fiscal policy (variations in taxes and spending) to maintain full employment, while keeping short term interest rates close to some ‘normal’ (or expected) level. Eel. Monetary policy was largely bypassed as a tool of demand-management.
The government forecast real GAP for the following year by forecasting year on movement of its expenditure components: consumption, fixed capital formation, stock building spending, and net exports. Budget deficits then adjusted to maintain full employment. There was no explicit modeling of expectations, though attention was paid to the issue of ‘confidence’. The prevalent view was that the confidence of the cuisines community was best maintained by a commitment to full employment. It was different with the balance of payments.
With sterling convertible into foreign currencies at a fixed exchange rate, governments also needed to retain confidence of non-resident holders of sterling, so the two requirements of confidence might pull in different directions. ‘Stop-Go’ was the result. Stop-Go not withstanding, fiscal activism proved highly successful, aided by the long post-war boom. The budget remained in surplus with current account revenues exceeding expenditure and with borrowing mostly stricter to finance public investment not covered by current-account surpluses.
Chancellors from Crisps to Macmillan were even tempted to extend this-above-the-line surplus to an overall surplus by covering capital expenditure below the line from revenue yet this was not achieved 1 . Nonetheless, the public-sector borrowing requirement (ESP.) fell from an average of 7. 5% of GAP (1952-1959) to 6. 6% of GAP (1960-1969). The national debt-to-income ratio fell from 3:1 in 1950 to 0. 7:1 in 19702. Unemployment was consistently below 2. 5% and inflation was low. Ill. THE RISE AND FALL OF PHILLIPS CURVE KEYNESIAN The post-war problem turned out to be not unemployment but inflation.
With full capacity utilization, whether generated by Keynesian policy or by benign world conditions, there was always going to be pressure on prices. So the attention of Keynesian policymakers was increasingly turned to fighting inflation, using both fiscal and monetary tools. In this they were also successful for a time. But from the late asses, inflation started to creep up; and the unemployment cost of restraining it started to rise: we enter the era of ‘stagflation’. The underlying theoretical question was: what caused inflation? Was it excess demand or ‘cost-push’?
There was no single Keynesian answer to this question. Some Keynesian economists argued that labor market was like any other, with price being determined by the balance between supply and demand. A reduction in the demand for labor would lower its price. Deflation would slow the rise of nominal wages, and hence a rise in the general price level. The question of course was how much deflation would be needed for stable prices? This was not an easy case for Keynesian to argue. Given their belief in sticky nominal wages, the unemployment cost might prove very high.
Most Keynesian economists were more comfortable with the ‘cost push’ theory of inflation: unions pushing up wages ahead of productivity. Prices rose because business managements raised them; managements raised prices because their costs had risen; costs rose owing to pay increases; and pay increased because otherwise unions would come out on strike. Higher unemployment would not stop them because most of the unemployed could not do the strikers’ jobs. In fact, cost-push could occur at levels well below full employment.
Short of bringing back mass unemployment, deflating demand would not stop inflation. What was required was a compact with the unions to restrain pay push: incomes policies. Anti-inflation policy in the 1 sass and asses wobbled between fiscal and monetary measures to restrain demand and attempts to reach pay deals with the unions. The Keynesian were rescued from this dilemma by the econometric work Of A. W. Phillips. In 1 958, A. W. Phillips published a famous article which claimed to demonstrate a well-determined relationship between the unemployment rate and the rate of wage increases.
The Phillips Curve implied that there was a stable trade-off between unemployment and inflation. The prize was price stability with a small increase in unemployment, way short of the depression. More generally, policy-makers were supposed to have a ‘menu of choice’ between different rates of inflation and unemployment. SLIDE 3. ORIGINAL PHILLIPS CURVE The Keynesian policy of demand-management unraveled with the attack on the Phillips Curve by Milton Friedman of Chicago University. In a single lecture in 1 968, he demolished Phillips Curve Keynesian and started the monetarist counter-revolution.
Adaptive Expectations Friedman restated the pre-Keynesian idea that there was a unique equilibrium rate of unemployment which he called the ‘natural rate’. Inflation was caused by government attempts to reduce unemployment below the natural rate by increasing the amount of money in the economy. Friedman accepted that there was a trade-off between inflation and unemployment, but that it was temporary, and existed only because workers were fooled into accepting lower real wages than they wanted by not taking into account the rise in prices.
But if government repeatedly resorted to monetary expansion (for example by running budget deficits) in order to educe unemployment below its ‘natural’ rate, this ‘money illusion’ would disappear and workers would put in increased wage demands to match the now expected rise in prices. In short, after a time workers developed inflationary expectations: they built the expected inflation into their wage bargaining. One could not use the Phillips Curve to control inflation in the long run since the Curve itself shifted due to the level of inflation rising. SLIDE 4.
FRIEDMAN’S EXPECTATIONS AUGMENTED PHILLIPS CURVE SLIDE 5. One simple version of adaptive expectations is stated in the following equation, where pee is the next year’s rate of inflation that is currently expected; p-Eel is this year’s rate of inflation that was expected last year; and p is this year’s actual rate of inflation: where is between O and 1. This says that current expectations of future inflation reflect past expectations and an “error-adjustment” term, in which current expectations are raised (or lowered) according to the gap between actual inflation and previous expectations.
This error-adjustment is also called “partial adjustment. ” Friedman’s work had huge anti-Keynesian policy implications. The five main Ones Were: First, macro-policy can influence nominal, but not real variables: the price level, not the employment or output level. Second, Friedman re-stated the Quantity Theory of Money, the theory that prices (or nominal incomes) change proportionally with the quantity of money. Conversely, fiscal ‘fine tuning’ operates with ‘long and variable lags’: it is liable to land the economy in the wrong place at the wrong time.
Consequently, such stabilization as was needed is much better done by monetary policy than fiscal policy. It lies within the power of the central bank, but not the Treasury, to keep nominal income stable. Provided the government kept money supply growing in line with productivity there would be no inflation, and economies would normally be at their ‘natural rate’ of unemployment. Third, Friedman argued that ‘inflation was always and only a monetary phenomenon’.
It was the total money supply in the economy which determined the general price level; cost pressures were not independent sources of inflation; they had to be validated by an accommodating monetary policy to be able to get away with a mark-up based price determination strategy; Fourth, Friedman’s permanent income hypothesis -dating from the early 9505 -suggested that it is households’ average long-run income (permanent income) that is likely to determine total demand for consumer spending, rather than fluctuation in their current disposable income, as suggested by the Keynesian consumption function.
The reason for this is that agents Want smooth consumption paths. This implied that the degree of self-stabilization of the economy was greater than Keynes suggested, and that consequently multipliers were smaller. Keynesian tried to fight the monetarist onslaught by strengthening Keynesian micro-foundations, especially of observed nominal rigidities. They plopped models with ;menu costs’, ‘insider-outsider’ labor markets, ‘asymmetric information’. These kept the door open for policy interventions to sustain aggregate demand. Nevertheless, Friedman’s impact on macro-policy was swift and decisive.
SLIDE 6 ‘We used to think that you could spend your way out of a recession, and increase employment by cutting taxes and boosting Government spending. Tell you in all candor that that option no longer exists, and that in so far as it ever did exist, it only worked on each occasion since the war by injecting a bigger dose of inflation into the economy, followed by a higher level of employment as the next step’. Prime Minister James Callaghan (1976), Leader’s speech, Blackball ‘The conquest of inflation should be the objective of macroeconomic policy.
And the creation Of conditions conducive to growth and employment should be the objective of microeconomic policy’. Chancellor of Exchequer Engel Lawson (1 984), Mass Lecture Discretionary demand-management was out; balanced budgets were back. The unemployment target was replaced by an inflation target. The ;natural’ rate of unemployment was to be lowered by supply-side policies, which included legislative curbs on trade unions. V. RATIONAL EXPECTATIONS AND THE NEW CLASSICAL ECONOMICS With rational expectations we enter the world of New Classical Economics. RE is the ‘radical wing of monetarism… Est. known for the startling policy conclusion … that macro-economic policies, both monetary and fiscal, are ineffective, even in the short-run’4. Rational expectations first appeared in the economic theory literature in a famous article by J. Mouth in 1961, but only filtered through to policy discussion in the early 1 sass with the work of Robert Lucas and Thomas Sergeant on business cycles, and Eugene Fame on financial markets. The Lucas critique Of adaptive expectations (1976) put paid to the idea Of an exploitable trade-off between employment and inflation.
Friedman’s adaptive expectations rely on gradual adjustment of expectations to the experienced behavior of a variable. But our knowledge includes not just what we have experienced but current pronouncements of public authorities and theoretical knowledge of aggregate relationships. For example, the Minister of Finance announces that he will increase money supply by 10% a year to stimulate employment. STEM tells us that an increase in the money supply will ease prices proportionately. So it is rational to expect inflation to be a year.
All nominal values -interest rates, wage rates- are instantly adjusted to the expected rate of inflation. There is not even a brief interval of higher employment. Friedman’s distinction between a Keynesian short run in which agents can be fooled and a Classical long run in which they know what to expect disappears. Adaptive behavior is a description of irrational behavior if agents know what to expect already. Notice though that in this example, rational expectations is defined as belief in the STEM.
SLIDE 7 Expectations, since they are informed predictions of future events are essentially the same as the predictions of the relevant economic theory… Expectations of firms (or more generally, the subjective probability distribution of outcomes) tend to be distributed for the same information set, about the prediction Of the theory (or the ‘objective’ probability distribution Of outcomes)’ (G. K Shaw (1 984), 56) Formally, the rational expectations hypothesis (ERE) says that agents optimally utilities all available information about the economy and policy to construct their expectations.
As such, such they have ‘rational’ expectations. They are also rational in that they use their expectations to maximize their utility or profits. This does not imply that agents never make mistakes; agents may make mistakes on occasion. However, all that is there to be learnt has already been learnt, mistakes are assumed to be random, so that agents are correct on average. Agents learn the true value of parameters through repeated application of Bases’ theorem. Eel they turn their subjective bets into objective probability distributions.
An equivalent statement is that agents “behave in says consistent with the models that predict how they will behave”6. Since the models contain all the available information, ii. They are rational expectations models, following the model minimizes the possibility of making expectation errors. At the core of the rational expectations hypothesis is the assumption that the model of the economy used by individuals in making their forecasts is the correct one -that is, that the economy behaves in a way predicted by the model.
The math is simplified by the device of the Representative Agent, the sum of all agents, possessed of identical information and utility preferences. This micro-economic device means that the framework can be used to analyses the impact of policies on aggregate welfare, as welfare is the utility of the agents. The implication of the ERE is that outcomes will not differ systematically from what people expect them to be. If we take the price level, for instance, we can write: SLIDE 8 This says that the price level will only differ from the expectation if there is a surprise.
So ex ante, the price anticipated is equal to the expectation. [E[P] is the rational expectation based on all information up to date; is the error ERM, which has an expected value of zero, and is independent of the expectation. With rational expectations the Phillips Curve is vertical in the short-run and in the long-run. SLIDE 9. THE SERGEANT-LUCAS PHILLIPS CURVE. With rational expectations, government action can affect real variables only by surprise. Otherwise they will be fully anticipated. This rules out any fiscal or monetary intervention designed to improve an existing equilibrium.
More generally ‘any portion Of policy that is a response to publicly available information -such as the unemployment rate or the index of leading indicators -is irrelevant to the real economy’ 7. Policy can influence real variables only by using information not known to the public. The Efficient Market Hypothesis The application of rational expectations to financial markets is known as the “Efficient Market Hypothesis” (MME), made popular by Eugene Fame (1970, 1976). The MME postulates that shares are always correctly priced on average because they adjust instantaneously and accurately to any newly released information.
In the words of Fame, “l take the market efficiency hypothesis to be the simple statement that security prices fully reflect all available information” 8. So prices can’t be wrong because if they were, someone would seek to profit from the error and correct it. It follows that according to the efficient market hypothesis, it is impossible to consistently achieve returns in excess of average market returns (beat the market). In an RE joke, two economists spot a $10 bill on the ground. One stoops to pick it up, whereupon the other interjects, ‘Don’t.
If it were really $1 0, it wouldn’t be there anymore. ” The efficient market hypothesis is the modern manifestation of Adam Smith’s ‘invisible hand’. Increased regulation can only aka markets less efficient because regulators have less information than those engaged in the market, risking their own money. There are different versions of the efficient market hypothesis. In its ‘weak’ form, investors make predictions about current prices only using historical information about past prices (like in adaptive expectations).
In its ‘semi-strong’ form, investors take into account all publicly available information, including past-prices. (This is the most ‘accurate’ and the closest to rational expectations). In its ‘strong’ form, investors take into account all information that can possibly be known, including insider information. Rational expectations models rely heavily on math. Lucas defined expectations as the mean Of a distribution of a random variable. The greater the number of observations of a random variable, the more likely it is to have a bell shaped or Normal distribution.
The mean of the distribution, in ordinary parlance the average of the observations, is called the Expectation of the distribution. In the bell-shaped distribution, it coincides with the peak of the bell. Those who are supposed to hold Rational Expectations (ii all of us) are assumed to know how the systematic parts of he model determine a price. We use that knowledge to generate our prediction. This will be correct except for random influences. We can assume that such random events will also adhere to the bell-shaped distribution and that their mean/expectation will be zero.
Thus the systematic or deterministic prediction based on theory is always correct. Errors have zero expectation. The tendency of the MME, as is readily seen, is to rule out, or minimize, the possibility Of bubbles -and therefore crashes; more generally to rule out the possibility of crises being generated within the financial system: historically he most important source of crises. This being so, policy did not have to pay much attention to banks. Following the acceptance of the MME, the financial system was extensively De-regulated.
Real Business cycle DOGS DOGS modeling takes root in New Classical macroeconomics, where the works of Lucas (1975), Jutland and Prescott (1982), and Long and Peoples (1983) were most prominent. The earlier DOGS models were pure real business cycle (RIB) models. ii models that attempted to explain business cycles in terms of real productivity or consumption shocks, abstracting from money. The logic behind RIB models is clear. If money cannot affect real variables, the source of any disturbance to the real economy must be non-monetary.
If we are all modeled as having rational expectations, business fluctuations must be caused by ‘real’ and ‘unanticipated’ ‘shocks’. (Notice the use of word ‘shock’). These shocks make the economy dynamic and stochastic. Unemployment is explained in these models by rational adjustments by workers of their work/leisure trade off to shifts in productivity. This is a fancy way of saying that there is never any unemployment. As a result of continuously re-optimizing agents, economies in DOGS models re always in some form of equilibrium, whether in the short run or long run.
The economy always starts from an equilibrium position, and even when there is a shock, it immediately jumps onto an equilibrium time path – the saddle path. So the economy never finds itself in a position of disequilibrium. SLIDE 10 ‘The model provides an example of an economy where real shocks drive output movements. Because the economy is Wallabies, the movements are the optimal response to the shocks. Thus, contrary to the conventional wisdom about macroeconomic fluctuations, here fluctuations do not reflect NY market failures, and government interventions to mitigate them can only reduce welfare.
In short, the implication of real-business cycle models, in their strongest form, is that observed aggregate output movements represent the time-varying Parent optimum’. (Roomer (2011 ) “Advanced Macroeconomics”, 204) Translated into English: depressions are optimal; any attempt to mitigate them will only make things worse. Later came the New Keynesian who preserved the basic framework of the New Classical RIB/DOGS models, but added ‘market frictions’, like monopolistic competition and nominal rigidities, to make the models more applicable to the real world. Critiques: 1 .
The fundamental criticism is that this whole class of New Classical models carries an intellectual theorem -that agents are rational optimizers – to an extreme and absurd conclusion. By postulating complete information and complete markets, ii. By abolishing Keynesian or Knighting uncertainty, they cut off enquiry into what might be rational behavior under uncertainty -such as ‘herd behavior’. They also exclude irrational expectations. Behavioral economics only really took off after the crisis. 2. The aim of New Classical economics was to unify macro and micro by giving macro-economic secure micro-foundations.
Macroeconomic models should be based on optimization by firms and consumers. But New Classical models are not well grounded in micro-economics since their account of human behavior is seriously incomplete. 3. Ay defining rational as the mean of a random distribution, the New Classical models rule out as too exceptional to worry about ‘fat tails’ – that is extreme events with disproportionately large consequences. 4. The vast majority of DOGS models utilities log-landslides utility functions which eliminate the possibility of multiple equilibrium. 1 5. New Classical models have no place for money, and therefore for money hoarding, which depends on uncertainty. In pure DOGS models there is no financial sector. DOGS models depend on what Goodhearted calls the ‘transversally condition’, which says that “by the end of the day, or when the model stops, all agents shall have repaid all their debts, including all the interest owed, with certainty. In other words, when a person dies he/she has zero assets left’ 12. Defaults cannot happen. This is another kind of logical madness.