Credibility and longrun reform Credibility and expectations Changing

Credibility and long-run reform Credibility and expectations Changing the long run

Credibility and long-run reform ¢ ¢ Last week we saw how stabilisation policies can be used to try an keep the economy at the long run macroeconomic equilibrium and minimise the impact of transitory shocks This week we examine how policy can attempt to modify the long run macroeconomic equilibrium l l We will not be examining economic growth just yet (week 12) Rather, we will be examining the conditions under which policy can influence the long run equilibrium of the AS-AD model (i. e. structural reforms)

Credibility and long-run reform ¢ The main argument we will focus on is the credibility of economic policy l l ¢ ¢ This has become central to policy-making and is the result of the debate on expectations. Because the Phillips and AS curves depend on expectation, their location can shift just by shifting people’s beliefs! Such “self-fulfilling prophecies”, will be seen in all the examples of policies. It can be a useful policy tool, but is double edged!

Credibility and long-run reform Expectations and the issue of credibility Reducing long run inflation Influencing long run unemployment and potential output

Expectations and the issue of credibility ¢ ¢ ¢ The first historical application of expectations is the debate on expected inflation mentioned in week 8 This debate occurred because central banks were trying to reduce the relatively high inflation levels of the 1970’s The main consequence was the introduction of the notion of credibility as a main aspect of how central banks carried out monetary policy l These issues will be clarified in the next section

Expectations and the issue of credibility ¢ After this, the “rational expectations revolution” also attempted to integrate agent expectations into aggregate demand (not just aggregate supply) ¢ ¢ Permanent income theory (Friedman) : Consumption is not a function of current income Y, but of permanent income, which is the expected average value of Y. Life cycle theory (Modigliani) : Similar, but takes into account the fact that the income of agents varies during their life, with alternating periods of saving (active life) and dissaving (university, retirement, etc. ). Ricardian equivalence : if the government increases its spending, (∆G>0), agents will expect a future increase in taxes (∆T>0) to pay back the debt and will increase their savings (fall in the mpc). The net effect on output is unclear (higher G, lower multiplier) ¢

Expectations and the issue of credibility ¢ ¢ The central consequence of the « Lucas critique » in terms of economic policy have been : A redefinition of the way policies are carried out : Credibility is important l l ¢ The expectations of agents have a very large impact on the final outcome of policies The objective of a policy is not only to modify the value of an economic variable, (π, U, C, I, etc. ), but also to modify the expectations on these variables. The modifications in the way economic policies are carried out in the 80’s-90’s (particularly monetary policy) l l Independence of central banks More transparency in decision-making (ex. publication of minutes)

Credibility and long-run reform Expectations and the issue of credibility Reducing long run inflation Influencing long run unemployment and potential output

Reducing long run inflation ¢ High inflation is a problem in macroeconomics. . . l ¢ In micro we saw that in theory, pure inflation is not a problem: all prices go up by the same amount. So why is it a problem in macro? l l l Prices of goods and factors don’t go up at the same speed (wages are agreed by contract) Inflation distorts production incentives (see Zimbabwe) Also, remember the Fisher equation:

Reducing long run inflation ¢ Unexpected inflation transfers wealth from lenders to borrowers (Why unexpected? ) l Because borrowers pay lenders a lower real l ¢ interest rate than agreed initially If the unexpected inflation is high enough real rates can even be negative. . . In particular, there can be an incentive for States to generate unexpected inflation: it reduces the real value of their debt. . . l This is called the inflation tax

Reducing long run inflation Volatile inflation Difficult to fomulate expectations Persistent/stable inflation French inflation (1902 -2007)

Reducing long run inflation ¢ The central bank wants to reduce inflation permanently from π h to π l The bank carries out the policy, and reduces money supply What happens if the policy is not credible ? πh The CB carries out its deflationary policy and unemployment increases πl Agents do not revise their expectations, and the Phillips curve does not shift. Sacrifice The economy ends up where it started…

Reducing long run inflation ¢ Imagine now that the agents believe the policy to be credible Assumption : Expectations adjust slowly : The CB carries out its deflationary policy and unemployment increases πh Then agents update their expectations πl Assumption : Expectations adjust instantaneously (Lucas) : Sacrifice The Phillips curve adjusts at the same time as the CB carries out its policy : No sacrifice! (remember the vertical Phillips curve)

Reducing long run inflation ¢ This shows that it is also important to take into account the speed of adjustment of expectations when the deflationary policy is carried out Even with rational expectations, it can be some time before variables adjust (for exemple, nominal rigidities on wages, à la Taylor) If the policy is carried faster than the rate of adjustment of expected variables, the initial movement will be along the SR curve: the sacrifice is high πh πl Ideally, you want a policy that is applied at the same rate as the expected variables adjust: Sacrifice The downwards shift of the Phillips curve compensates exactly the downward movement on the curve

Reducing long run inflation ¢ Therefore, a disinflationary policy that wants to permanently reduce inflation and minimise the social cost of the policy (employment sacrifice), you need to account for: l l l The interaction between agents and policy-makers (Central bank and government) in terms of credibility. The expectations of agents (they may be heterogeneous!) The existence of nominal rigidities that will affect the agents’ capacity to adjust their expectations through time.

Reducing long run inflation ¢ Empirical evidence : l Disinflation policies always lead to an increase in unemployment: even taking into account expectations, it does not seem possible to avoid the short-run Phillips curve (no “silver bullet”) l l However; faster disinflations have a lower sacrifice ratios, slower ones have a higher sacrifice ratio: it seems the effect of credibility on expectations is validated. The faster wages are re-negotiated (or the shorter the contracts), the smaller the sacrifice ratio: nominal rigidities are important in slowing down the adjustment of expectations.

Reducing long run inflation ¢ ¢ What about the opposite case: credibly committing to inflation ? This situation is outlined by Paul Krugman (NYT articles) l l ¢ Historical case of Japan in the 90’s (the “lost decade”) Is it applicable to the current case? This is the problem: how does a central bank historically committed to low inflation get out of a deflation situation? l It’s the opposite case to the previous one !

Reducing long run inflation ¢ Deflation can be an even bigger problem than inflation ! l ¢ With deflation, money becomes an asset Remember the Fisher equation l l l The nominal interest on money is zero, but with deflation, the real interest is positive ! This is because agents expect prices to be lower in the future, so hold up their spending plans Money is now a good investment !

Reducing long run inflation ¢ Coupled with a preference for liquidity this means that money will be hoarded for itself. l l ¢ Typically, a deflation prolongs the duration of a depression through a liquidity trap situation l ¢ If nominal rates are low (the case in a recession), agents are indifferent to money and assets. This will depress aggregate demand even more! It is therefore important to find a way out of them. How does a central bank increase prices ? l Print money! But will it work? Not necessarily!

Reducing long run inflation The economy is in a deep recession/depression : un is high, inflation is negative (we’ll see in the next section how this may occur) The CB wants to push inflation from π - to π + and prints lots of cash ¢ ¢ What happens if agents don’t think the bank wants to permanently increase inflation? π+ 0 π- Agents do not revise their expectations, and the Phillips curve does not shift. The economy ends up where it started…

Reducing long run inflation Only if the central bank can credibly commit to higher inflation in the future will the policy be successful! ¢ Agents will revise their expectations, and the Phillips curve will shift upwards, taking the economy out of deflation π + 0 π- According to Paul Krugman, it is just as hard for the ECB, the Bo. E or the FED to commit credibly to inflation as it is for Zimbabwe to commit to low inflation. . . Which is why “quantitative easing” policies are being introduced, as well as the “ZIRP”

Credibility and long-run reform Expectations and the issue of credibility Reducing long run inflation Influencing long run unemployment and potential output

Long run unemployment and output ¢ As outlined previously, the Phillips curve and the AS-AD model (which includes the PC on the supply side) tend to return to a “natural” long run equilibrium l l ¢ These are vertical, and are not affected by prices in the long run. But they are affected by other variables How can macro policy influence these? l l Reduce the natural rate of unemployment Increase the natural level of output (discounting growth for the moment)

Long run unemployment and output ¢ ¢ A first element is the “structural reform” policies on the labour market. Remember the WS-PS model: The structural rate of unemployment is a function of l l The mark-up on the goods markets μ The labour market conditions z (unemployment benefits, etc. )

Long run unemployment and output Structural rate of unemployment un (long run) 1 st aspect: reducing the mark-up on the goods market. μ‘ < μ Real Wage 2 nd aspect: reducing the market conditions parameter z (smaller unemployment benefits, etc. ) B PS’ A PS WS un un WS’ Unemployment rate u

Long run unemployment and output ¢ Such policies have been put in place in many countries. l l l ¢ ¢ They are not necessarily ultra-liberal policies For example, reducing “z” is not just cutting benefits. Benefits can be made conditional to re-training, with extra funding allocated (Denmark) In many cases, they have lead to more flexible labour markets, with lower rates on unemployment But right now, they’re not the “best” policies

Long run unemployment and output ¢ ¢ A more important aspect is the question of the endogeneity of the “natural” rates In other words, what if short run shocks could have long-run effects ? l ¢ Then short run policy becomes important in the long run! This is linked to the idea of persistence and hysteresis l l Persistence : a short-run shock has long-run effects Hysteresis : when a phenomenon continues after its cause has been removed

Long run unemployment and output ¢ Illustration with a labour market example: l l ¢ When unemployed, the probability of finding a job decreases with the length of the unemployment spell This is because your skills/employability deteriorates with time A negative shock to the economy l l l Will increase the short-run unemployment If nothing is done (or too little), then the “natural” rate will increase, and the shock will be persistent. The “natural” rate follows an error-correcting process:

The AS-AD equilibrium π LRAS 2 A large negative demand shock shifts the AD curve to the left, which reduces output pushes the economy into deflation LRAS SRAS The government does not intervene in time, or too little. A π1 π3 0 In the mean time, the LRAS shifts left : C The “natural” rate of unemployment increases because of hysteresis π2 AD B If the hysteresis is large enough, the economy can end up in structural depression (Japan’s lost decade) AD 3 AD 2 Y 1 Y 2 n Yn Y