ch_8_classical_models_and_monetary_policy.pdf

Chapter 8 Classical models and monetary policy 8.1 Introduction In Chapter 5 we introduced the classical model where money was neutral. Real outcomes such as investment decisions and output were determined by real factors; tastes and technology, and the money supply only determined the price level...

Chapter 8 Classical models and monetary policy 8.1 Introduction In Chapter 5 we introduced the classical model where money was neutral. Real outcomes such as investment decisions and output were determined by real factors; tastes and technology, and the money supply only determined the price level. In this chapter we will examine classical models in more detail, exploring the possibilities that monetary policy can have real effects, at least in the short run. However, in order for us to do this, we must first consider the benchmark case of the classical model and we do this by considering a model based on ‘reduced form’ macro equations.1 A ‘simple’ model is introduced and solved in the appendix but you are not expected to develop, or be able to solve, these models in the examination. They are purely left for the more interested reader who wishes to further understand how RBC theory works. 8.2 Aims This chapter introduces first major macro model without nominal ‘frictions’ that will serve as a benchmark model. We then study cases with real frictions in the classical model (in the form of asymmetric information, the requirement to hold cash before you could buy, and limited participation in financial markets) all lead to money having real effects. 8.3 Learning outcomes By the end of this chapter, and having completed the Essential reading and activities, you should be able to: describe how, in the classical economy, output is determined by the factor markets describe and discuss the effects of monetary policy, with reference to the effects of money on nominal variables such as prices and nominal wages, and on real variables such as employment describe what business cycle models are and what they try to do list and explain the workings behind, the flexible price models where money has real effects. 1 I.e. equations we assume hold in this particular instance. 101 8. Classical models and monetary policy 8.4 Reading advice Before working through this chapter, it is vital that you have a thorough understanding of the macroeconomic classical models. Background reading can be found in the textbooks of Mankiw (2002) and Branson (1989) among others. In this chapter, we discuss the Classical model in more detail, analysing the effects of monetary policy. Once you have worked through the chapter, you can read the articles in the reading list; those of Long and Plosser (1983) and Plosser (1989) are probably the most readable. The chapters in Hoover and in Hargreaves Heap are also very useful. 8.5 Essential reading Hargreaves Heap, S.P. The New Keynesian Macroeconomics: Time, Belief and Social Independence. (Aldershot: Edward Elgar Publishing, 1992) Chapter 4. Hoover, K.D. The New Classical Macroeconomics. (Oxford: Blackwell, 1988) Chapter 3. Long, J. and C. Plosser ‘Real business cycles’, Journal of Political Economy 91(1) 1983, pp.39–69. Plosser, C. ‘Understanding real business cycles’, Journal of Economic Perspectives 3(3) 1989, pp.51–77. 8.6 Further reading Books Branson, W.H. Macroeconomic Theory and Policy. (New York; London: Harper and Row, 1989). Mankiw, N.G. Macroeconomics. (New York: Worth Publishers, 2002). Walsh, C.E. Monetary Theory and Policy. (Cambridge, Mass.: MIT Press, 2003) Chapter 1. Journal articles King, R.G. and C. Plosser ‘Money, credit and prices in a real business cycle’, American Economic Review 74(3) 1984, pp.363–80. Kydland, F.E. and E.C. Prescott ‘Business cycles: real facts and a monetary myth’, Federal Reserve Bank of Minneapolis Quarterly Review 14(2) 1990, p.3. Lucas, R.E. Jr. ‘Some international evidence on output-inflation trade-offs’, American Economic Review 66(5) 1976, p.985. Lucas, R.E. Jr. ‘Nobel lecture: monetary neutrality’, Journal of Political Economy 104(3) 1996, pp.661–82. 102 8.7. The classical model revisited 8.7 The classical model revisited As should be clear by now, in the classical model all markets do clear instantaneously such that output is at its natural rate, resulting in involuntary unemployment being equal to zero. Note that the labour market variable is hours worked, not the level of employment, thus when individuals decide to work they always find work at a given wage rate. Prices are perfectly flexible and move instantly to clear both goods markets and factor markets (the prices in the labour and capital factor markets are wages and real rates, respectively). We shall assume that output in the economy is determined only by labour and capital. y = f (k, l) (8.1) where k is the amount of capital, which for the time being we will assume to be fixed, and l is the amount of labour. Also, f1 (the derivative of the production function with respect to the first element) > 0, f2 > 0, f11 < 0, f22 < 0 and f12 > 0. That is, factors are complementary in production and they have positive but decreasing marginal productivities. The demand for labour by firms will depend negatively on the real wage that they have to pay, W/P . If labour is more expensive, firms will demand less of it. Firms demand labour up to the point where the marginal cost of production, the real wage, equals the marginal benefit from production, the marginal product of labour. Due to the properties of the production function, namely diminishing marginal returns, a high level of labour is associated with a low marginal product so firms are only willing to pay a low wage at that employment level. The supply of labour is positively related to the real wage since a high wage encourages individuals to work more hours and causes those who initially chose not to join the labour force, to participate in work. Labour market equilibrium is shown in Figure 8.1. If capital is fixed at k ∗ , then output is determined as y = f (k ∗ , l0 ). 8.8 The effect of monetary policy The effect of monetary policy can be best shown using the set of diagrams in Figure 8.2, below. The middle left diagram shows the labour market equilibrium. This is the same as Figure 8.1 only we have the nominal, as opposed to the real, wage on the vertical axis. This is because we are examining the effects of changing nominal variables, namely the nominal money supply. The bottom left diagram is the production function for a given level of capital, k ∗ , and shows the decreasing marginal returns to labour. The top right panel shows the standard IS–LM analysis and the middle right panel shows the aggregate demand and supply schedules. Note that the aggregate supply schedule is vertical in the classical model since markets always clear which allows output always to equal its natural/full employment rate, y 0 . Suppose the famous ‘helicopter drop’ analogy of monetary policy of Friedman. Imagine an economy, where authorities conduct monetary policy by sending off helicopters to drop cash over the country. Such an increase in the money supply, caused by a 103 8. Classical models and monetary policy Figure 8.1: Labour market equilibrium. ‘helicopter drop’ of newly printed money or by open market operations, whereby the monetary authorities buy government bonds in exchange for money, leads to an outward shift of the aggregate demand curve to AD1 .2 The monetary expansion also causes the LM curve to initially shift out to LM1 . However, the increase in demand has no effect on output since the AS schedule is vertical at y 0 , and simply causes the price level to increase from P0 to P1 . The increase in the price level reduces real money balances, causing the LM curve to shift back to its initial position, LM0 . Note that although the outward shift of the LM curve, caused by the initial monetary expansion, suggests at least a temporary increase in output, the price level increases immediately so that the LM curve effectively never changes position. The increase in the price level causes the labour demand schedule to shift out. Note that we have the nominal wage on the vertical axis. For any given nominal wage, a higher price implies a lower real wage and hence an increase in labour demand. The price rise also causes the labour supply schedule to shift to the left. Again, for any given nominal wage, the price rise implies a fall in the real wage so individuals supply less labour. In equilibrium the nominal wage increases to the point where the real wage remains at the initial level. Real wages and real money balances are left unchanged in equilibrium; and output, labour, and so on are all unaffected. All the change in the money supply does is to change other nominal variables, nominal wages and prices, one-for-one. Money, as expected, is neutral. 2 Why a change in nominal money balances affects the AD schedule is documented in all macroeconomic texts. See for example Mankiw (2002). 104 8.8. The effect of monetary policy Figure 8.2: 105 8. Classical models and monetary policy 8.9 Real business cycle theory Real business cycle (RBC) theory extends the classical model in that it derives the ad hoc macroeconomic equations, represented in Figure 8.2, from the optimising behaviour of agents. Money still turns out to be neutral in the basic set-up but business cycles are argued to be the natural and efficient responses of the economy to changes in the available production technology. It distinguishes between nominal shocks (shocks to money supply or money demand) which only affect the LM curve and real shocks (shocks to the production function, real government spending or to savings and consumption decisions) which affect only the IS or AS curves. According to RBC theory, real shocks to the economy are the primary cause of business cycles. The theory focuses on shocks to the production function: supply shocks or productivity shocks. Such shocks can be the result of: the development of new products or production methods changes in the quality of labour or capital changes in the availability of raw materials unusually good or bad weather and changes in government regulations affecting production. Economic booms result from beneficial productivity shocks and recessions are caused by adverse productivity shocks. Consider a temporary adverse supply shock. The marginal product of labour falls as labour becomes less productive, which reduces the demand for labour. Both the real wage and the equilibrium level of employment fall. The latter causes a fall in output (a recession). 8.10 Business cycle facts and RBC theory RBC theory does a good job at explaining a number of the stylised facts of the macroeconomy and of the business cycle. RBC models explain the procyclical nature of employment, labour productivity and the real wage and also explain why investment is more volatile than consumption.3 This last point is explained by the fact that agents, being riskaverse, prefer to smooth consumption streams over time. Any fluctuations in income due to productivity shocks will be absorbed into the saving/investment decisions of individuals, resulting in smooth consumption and volatile investment. However, despite explaining these features of the business cycle, it is, in reality, very difficult to identify the productivity shocks that cause the cyclical fluctuations. Also, in order to explain the large cyclical fluctuations in hours worked and the small cyclical fluctuations in the real wage observed in the real world, RBC theory needs a flatter aggregate labour supply curve than has been found by most studies. RBC theory also predicts that the price level will be countercyclical when the empirical evidence for real output and inflation is, as we saw in the ‘Stylised facts’ chapter, not conclusive. Also, 3 We define a variable as being procyclical (countercyclical) if it moves positively (negatively) with output. 106 8.11. Classical models with real effects of money RBC theory, being an extension of the classical model, suggests that money is neutral. However, the problem with claiming that money is neutral is that the money stock has been found to be a leading, procyclical variable. RBC theorists argue that the apparent relationship between money and output is one of reverse causality; anticipated changes in output lead to changes in the money supply in the same direction. Historical research as discussed in Chapter 7, however, suggests that money is not short-run neutral. 8.11 Classical models with real effects of money In the standard classical model analysed above, money has no real effects. As mentioned before, this is a pitfall of standard classical theory; money has been found to have real effects at least in the short run. However, with some modifications to the reference model, money can have real effects, even in models with perfectly flexible prices. These are discussed below. Lucas ‘misperceptions’ model In the standard classical model every individual knows what the money supply is and there is full and symmetric information about the state of the economy and about the state of each market. Relaxing this assumption will allow money to have real effects. Suppose an individual who produces goods in one market sees the price of his good increasing due to an increase in demand. If that was caused by a relative demand shock for his good (i.e. preferences changed so that the demand for his product increased relative to some other goods) then it will be optimal for him to increase production, essentially moving along his upward sloping marginal cost/supply curve. If, on the other hand, the price rise was caused by an economy-wide increase in prices as a result of the money supply increasing, then since nothing ‘real’ has changed, the demand for his good has not altered relative to the demand for other goods, then it will be optimal for him to leave output unchanged. In reality, it will be unlikely that he knows the cause of the increase in price and so he will opt for an intermediate strategy whereby he increases output by a little (not as much as if he knew the price rise was caused by a relative demand shock but more than if it was due to a monetary expansion, that being a zero output change). The Lucas aggregate supply function is then given by (8.2). yt = y ∗ + d(Pt − Et−1 [Pt ]) (8.2) where yt is aggregate supply, y ∗ is the full employment/market clearing level of output and Pt is the price level at date t. At date t − 1, individuals make an expectation of the price level in date t, denoted Et−1 [Pt ]. If the price level in date t is greater than expected, Pt − Et−1 [Pt ] > 0, agents believe this is caused by a relative demand shock in date t, or by an increase in the money supply that was unexpected. As a consequence, each individual/firm produces more, which causes aggregate supply to increase above y ∗ . Hence d is positive and the aggregate supply schedule is upward sloping rather than vertical. Also, an unexpected increase in the money supply can have real effects even though prices are perfectly flexible. People effectively ‘misperceive’ an increase in the price level as having been 107 8. Classical models and monetary policy caused by a relative demand shock, rather than an increase in the money supply. If the monetary expansion was perfectly anticipated rather than unexpected, then everyone would allow for this when making their expectation of Pt at date t − 1 and the monetary expansion would have no real effects.4 Even though monetary policy has real effects in this set-up, the effects only persist in the short run. In the long run, when people realise that the price rise is due to a monetary expansion, output will revert back to its full employment level. This is true not only of the Lucas misperceptions model but of all classical models that allow real effects of monetary policy. Cash in advance models A problem with the money in utility (MIU) model is that it assumes people obtain utility from something that is intrinsically useless. Adding money to the utility function is then an odd way of introducing money. The cash in advance (CIA) model, however, uses cleaner micro-foundations to explain the existence of money. It effectively tries to model the trading frictions, such as the lack of trust between payer and payee discussed in Chapter 2 that causes the trading process to necessitate a medium of exchange – cash. People need cash in order to buy goods and sellers of goods will accept nothing but cash when they trade. The trading round is assumed to be split into two periods. In the first period individuals receive an income and decide how much of their wealth to hold in cash and how much to hold in other forms such as bonds or capital that cannot be used as a medium of exchange. In the second period individuals buy and consume goods, facing the constraint that the nominal value of their consumption must not be more than the nominal money holdings they brought forward from the first period. Therefore, in addition to the budget constraint, individuals face a ‘cash in advance’ constraint: Pt Ct ≤ Mt−1 . (8.3) Dividing by Pt−1 and noting that Pt /Pt−1 = 1 + πt , this can be written as: (1 + πt )Ct ≤ mt−1 (8.4) where mt−1 = Mt−1 /Pt−1 , and is equal to real money balances at date t − 1. (8.4) implies that an increase in πt acts like an inflation tax; the consumers cannot purchase in real terms the same amount of goods. Since their utility depends on consumption and leisure, they will substitute (the now relatively more expensive) consumption goods for leisure. That is, they will decrease their supply of labour due to the increase in inflation. Thus employment, and therefore output, will decrease as a result of the increase in inflation. Limited participation models Monetary policy can have real effects if there are a limited number of agents participating in financial markets meaning some consumers may be restricted in their 4 See Chapter 9 and the section on the policy ineffectiveness proposition. 108 8.12. A reminder of your learning outcomes access to the banking system and therefore cannot save via interest bearing assets. If the monetary authorities decided to increase the money stock, such policy actions would be made through banks and other financial intermediaries with whom the open market operations were conducted. Now faced with a glut of liquid assets, banks wish to lend out some of these in order to maintain their desired reserve ratio. The increase in the supply of loans will cause the interest rate charged on these loans to fall (the liquidity effect) and since firms borrow from financial intermediaries to finance investment projects, this makes investment cheaper. Cheaper investment causes investment to increase, which therefore causes output, and employment if labour and capital are complementary inputs, to increase. 8.12 A reminder of your learning outcomes By the end of this chapter, and having completed the Essential reading and activities, you should be able to: describe how, in the classical economy, output is determined by the factor markets describe and discuss the effects of monetary policy, with reference to the effects of money on nominal variables such as prices and nominal wages, and on real variables such as employment describe what business cycle models are and what they try to do list and explain the workings behind, the flexible price models where money has real effects. 8.13 Sample examination questions Section A Specify whether the following statement is true, false or uncertain. Explain your answer in a short paragraph. 1. ‘According to the stylised facts of the business cycle, real wages, employment and government spending are acyclical.’ Section B 2. Explain why money might be neutral in the long run but not in the short run. 3. Suppose the supply curve is given as in (8.2) and that aggregate demand is given by: y = 1000 + 3(M/P ). Suppose there has been no shock in the economy for some time and no changes in policy are expected in the near future. If M = 600, find y and P in terms of y ∗ . What happens when the monetary authorities announce (in advance) that M will increase to 650? What happens if this increase is entirely unexpected? In this final part, you do not need to solve for P and y. 109