If a model were to predict that prices and money are inversely related, that prediction would be evidence against that model.

Transcription

1 The Classical Model This lecture will begin by discussing macroeconomic models in general. This material is not covered in Froyen. We will then develop and discuss the Classical Model. Students should read Chapter 3 of Froyen. The final part of lecture will discuss the quantity theory of money. Students should read the beginning of Chapter 4 in Froyen. A model tries to explain a set of endogenous variables. The values of endogenous variables are part of the model s solution, they are not taken as given. Output and consumption are common endogenous variables in macroeconomic models. Exogenous variables are those whose values are taken as given and are not part of a model s solution. For example, government spending and taxes will be exogenous variables in the Keynesian model. You will simply be told their values and use those values to solve for the endogenous variables. A model need not explain why an exogenous variable takes a given value. As a model adds more endogenous variables (often by making an exogenous variable endogenous), it explains more about the economy and offers more policy recommendations. It also becomes more difficult mathematically and it may be harder to understand the results. While developing the Keynesian model, we will initially take interest rates and the price level as exogenous. As the course progresses, we will make these variables endogenous, yielding a better model, but one that will be more complicated. Before this course begins to develop specific macroeconomic models, it is useful to discuss how macroeconomic models work in general and what determines whether a model is any good. First, a pretty bad joke, allegedly told by the Nobel Prize winner Paul Samuelson: A physicist, a chemist and an economist are stranded on an island, with nothing to eat. A can of soup washes ashore. The physicist says, Lets smash the can open with a rock. The chemist says, Lets build a fire and heat the can first. The economist says, Lets assume that we have a can-opener... 1

2 The basis of model is the set of assumptions that define it. Good assumptions result in a good model while bad assumptions result in a bad model. There are two easy ways to judge an assumption: 1. Is it realistic? The aforementioned joke was funny (or at least supposed to be funny) because the economist made an assumption that was absurd. As a result, his ensuing suggestion was sure to be of little use. Realistic assumptions, however, usually lead to useful predictions/suggestions. The Keynesian Model is going to assume that households consumption is an increasing function of their disposable income. Because this is a realistic assumption, the model will make respectable predictions about consumption. 2. Is it easy to work with? Macroeconomists use models so that they can understand how a complicated world operates. A good assumption does not need to exactly match reality, instead it should explain a significant part of reality while being simple enough to actually use. Consider the previous assumption about consumption. Besides disposable income, many other things (interest rates, risk aversion, uncertainty, etc.) also explain part of consumption. Including everything in a consumption function, however, would make the model too complicated for this course. The assumption about consumption is a good one because it explains a lot about consumption (but not everything), while also being very easy to work with. After making a set of assumptions, macroeconomists use mathematics (this course will use both graphs and algebra) to show how the economy works. The Keynesian model, for example, predicts that if the government prints more money, the price level increases. Another way to judge a model is by how well its predictions match the actual data. Because the U.S. data does suggest that the money supply and prices are positively related, the previous prediction provides support for this model. If a model were to predict that prices and money are inversely related, that prediction would be evidence against that model. 2

3 Most models make many predictions. As a result, almost every model makes some predictions that do not match the data. Rather than rejecting a model based on a few erroneous predictions, it is more useful to ask: 1. Are its most important predictions supported by the data? 2. Are most of its predictions correct? A good model is one where both questions are answered yes. In this case, macroeconomists often refine the model to try to understand why some of its predictions are incorrect. Macroeconomists are constantly confronted with a tradeoff between making a model more realistic at the price of more complexity. How much complexity to accept depends on the purpose of the model. Suppose that an economist only cares about predicting future GDP and has little or no desire to understand why his model makes a given prediction. He will likely use a very complicated model. A theorist, however, typically wants to understand the inner workings of his model and will therefore likely choose less complexity. The Classical Model The first model that the course will discuss is the Classical Model. Before the Great Depression, macroeconomics was not a distinct field from microeconomics. The Classical model dominated what would then become known as macroeconomics. Many macroeconomics have long judged the realism of an assumption by whether it is based on solid microeconomic foundations. The Classical model was appealing because it consisted mostly of elementary supply and demand. Its derivation will likely remind you of Intermediate Microeconomics. The Classical Model is now obsolete. It is worth studying (briefly), however, because macroeconomists have made several popular efforts in recent decades to come up with a modern counterpart based on similar microeconomic foundations. These modern Neo-Classical models are far 3

4 more complex than the original but tend to have similar policy implications. The Classical Model is a model of short-run output determination. The endogenous variables are output and the wage paid to labor. Its starting point is the aggregate production function: Y = F ( K, N) Output is an unspecified function of two inputs, capital (K) and labor (N). The bar over K reflects the model s assumption that capital is exogenous and constant in the short run. By assumption, as N increases, so does Y. The marginal product of labor (MPN) is the change in output caused by a one unit increase in labor. By assumption, the MPN decreases as N increases. Graph (The Aggregate Production Function) Suppose that the economy consists of only one large firm (we can get the same results by instead assuming that the economy consists of many identical small firms) A profit maximizing firm must decide how much labor to employ. Labor receives a wage (W). Recall from Introductory Macroeconomics that a perfectly competitive firm sets its price equal to its marginal cost. The marginal cost of producing an extra unit of output is the amount of labor needed to produce that output (1/MPN) multiplied by the wage that labor must be paid. 4

5 P = W/MP N The real wage (W/P) equals the nominal wage (W) divided by the price level (P). This measures the actual purchasing power of the wage. Suppose that the price level doubles. The same wage now can buy only half as much as before. This change is captured in the real wage, but not the nominal wage. The previous equation shows that the real wage equals the marginal product of labor. Because the MPN decreases as employment (N) increases, the aggregate (economy wide) labor demand curve is downward sloping. Graph (Aggregate Labor Demand) Households must also decide how much labor to supply for each real wage. Households value both consumption (obtained by supplying more labor) and leisure (obtained by supplying less labor). Suppose that the real wage increases. There are two effects that influence households decision whether to supply more or less labor: 1. The price of leisure is the opportunity cost of lost wage income. As W/P increases, so does the 5

6 price of leisure. Workers will likely substitute away from leisure towards consumption by supplying more labor. This is an example of the substitution effect from Intermediate Microeconomics. 2. As W/P increases, so does the real income of households. They are now able to afford more of both consumption and leisure. This income effect causes households to buy more leisure by supplying less labor. Suppose, for example, that a person working 80 hours per week at minimum wage gets a raise to ten billion dollars per hour. You would expect that person to work less, not more. It is not obvious which effect dominates in theory. The data, however, clearly shows that the first effect dominates in the overall U.S. labor. Aggregate labor supply is therefore upward sloping. The Classical Labor market is essentially basic supply and demand. The supply curve (from households) is upward sloping and the demand curve (from firms) is downward sloping. Where aggregate labor demand and aggregate labor supply intersect, equilibrium occurs. This determines the equilibrium price (known in the labor market as the real wage) and the equilibrium quantity (known in the labor market as employment). Graph (Classical Labor Market) 6

7 We will now derive the aggregate supply curve for the Classical model. The AS curve is all combinations of P and Y where the labor market is in equilibrium. Later in the course, we will derive the AS curve for different models. The approach is always the same: change P while holding all other exogenous variables constant and plot the effect on Y. Suppose that price level increases from P 0 to P 1. The Classical Labor Market depends on the real wage (W/P). The LS and LD curves do not move in response to the change in P. W is an endogenous variable. If P increases by X percent, then W must also increase by X percent to keep the labor market in equilibrium. (Note: The model makes no assumption restricting W from changing. The Keynesian model will make such an assumption and will obtain very different results). W/P is unchanged, as is N. Because N is unchanged, so is Y. In the Classical Model, changes to P do not affect Y. The graph of the aggregate supply curve always has P on the vertical axis and Y on the horizontal axis. In the Classical Model, it is vertical. Graph (Classical AS Curve) The Classical Model was the dominant model in macroeconomics prior to the Great Depression. It makes a pair of stark policy predictions. First, because the AS curve is vertical, it is very hard for policy makers to affect Y. 7

8 The second prediction concerns unemployment. It is important to distinguish between voluntary unemployment and involuntary unemployment. Voluntary unemployment refers to workers who could find a job at the prevailing wage rate, but choose not to. Many students are examples of voluntary unemployment. Policy makers do not worry very much about voluntary unemployment. Involuntary unemployment refers to workers who want to work at the prevailing wage but are unable to do so. In the graph of the labor market, involuntary unemployment is a surplus of labor. It equals the space between labor supply and labor demand. Graph (Classical Labor Market) Notice that because supply equals demand, there is no involuntary unemployment in the Classical Model. All unemployment is voluntary. This is the second prediction of the model. During the Great Depression, the unemployment rate neared 25 percent, mostly involuntary. It became obvious to economists that the Classical Labor Market was flawed and the absence of involuntary unemployment was too severe an omission for the model to be taken seriously. The Great Depression therefore motivated the development of the Keynesian Model, a model that will allow for involuntary unemployment. 8

9 The Quantity Theory of Money The next part of this lecture discusses the quantity theory of money, an extension of the Classical Model. Students should read through page 65 of Froyen. Introductory Macroeconomics discussed the circular flow. Recall that the purpose of that graph was to demonstrate that one party s expenditure is another s income. National income therefore equals aggregate expenditures. The starting point for the quantity theory of money is the equation of exchange: P Y MV Note that this equation is an identity. P is the price level and Y is real GDP. PY is therefore a measure of nominal national income. M is the money supply. Later chapters will discuss how the Central Bank (the Federal Reserve in the United States) determines M. V is the velocity of money. The velocity of money equals the average number of times per period that each unit of money is spent. MV is therefore aggregate expenditures. The equation of exchange is an algebraic representation of the circular flow. The quantity theory of money makes two assumptions: 1. The velocity of money is constant in the short run. 2. Changes to P do not affect Y. The Classical Model may be cited as support for this assumption. Re-writing the equation of exchange: P = V M/Ȳ Suppose that the Fed changes the money supply by X percent. For the equation of exchange to hold, P must also change by X percent. The quantity theory states that changes in M only result in proportional changes in P. This result has monetary policy implications. The Central Bank cannot affect Y but it can affect inflation. It should grow the money supply slowly to keep 9

10 inflation low. This policy suggestion is a common feature of current Neo-Classical models (though not always older Neo-Classical models). The quantity theory of money may be thought of as part of the Classical model. Define aggregate demand as all pairs of P and Y where the equation of exchange holds. (The Keynesian Model will use a slightly different definition of aggregate demand). For a higher value of P, Y must decrease. Graph (Aggregate Demand) Now suppose that M increases. For the equation of exchange to hold, Y must be higher. This is true for any P. The AD curve shifts to the right. Notice that, in equilibrium, P increases and Y is unchanged, confirming the result from the Classical Model. Monetary policy does not affect Y in the Classical Model. Although lecture will not discuss 4.2 in Froyen, reading this section may help with your understanding of the model. Section 4.3 will be assigned during this course s discussion of supply side policies. 10