Introduction: Analytical structure of macroeconomics and monetary theory.
The historical field of macroeconomic theory is to explain the level and fluctuation of overall economic activities. Macroeconomists always pay special attention to the effects of different fiscal, monetary and monetary policies. Because J·M· Keynes's General Theory was published in 1936, and J·R· Hicks (1937) and others used mathematics to state his theory, the language of macroeconomic theory became simultaneous equations. Each relationship describes the whole national economy, not just a specific industry or department, but all interdependent equilibrium systems in this sense. These systems are usually not completely closed; They depend on external parameters, including tools controlled by decision makers. These models seek the exact qualitative and quantitative relationship between exogenous variables such as politics and economic consequences, sacrifice details and universality, and limit the number of variables and equations by adding economic subjects, commodities, assets and time.
The theoretical macroeconomic model of this or that brand is very influential. They guide the designers of econometric forecasting models. They have shaped the thinking of policy makers and their advisers about the way the world works. They influence the opinions of journalists, managers, teachers, housewives, politicians and voters. Almost everyone is thinking about the economy, trying to understand it, and has their own views on how to improve its performance. Anyone who does this will use a model, even if it is vague and informal.
(A) Keynes-Hicks unique model and another structure
Hicks' (1937) version of Keynes and Classical Theory has always been particularly influential, not only reaching the level of professional economics -LMts, but also influencing several generations of college students as a standard macro model of textbooks. Its simple equipment is intuitive knowledge that many of us have trained when facing policies and analyzing problems, no matter what more detailed methods we may use in further research. However, this structure has some shortcomings, which limit its usefulness and make it criticized. In this lecture, I want to describe another structure, which tries to repair some of its shortcomings. At the same time, I will argue that the main conclusion of Keynes-Hicks equipment is still good. The reconstruction work I want to summarize is my long-term work, and I will definitely absorb the previous work.
The main features that distinguish the proposed structure from the standard macro model are as follows:
The accuracy of time. The short-term determination model of macroeconomic activities must refer to a short period. This is a step in a dynamic sequence, not a repeated equilibrium when the economy is stable.
Track inventory. The main parts of this process are flow and stock, investment and capital, savings and wealth, specific forms of savings and asset stock. The excuse analysis is aimed at such a short time, so it is impossible for the stock to change significantly, and it is not always enough to ignore these relationships.
Several assets and returns. All non-monetary assets are traditionally summarized as single assets and general interest rates, so it is impossible to analyze some important political and institutional structures and events. In principle, my other structure can accommodate as many asset classes as possible, although the illustrative application described below only distinguishes four types of assets. Asset decomposition is very important for analyzing capital accumulation and government deficit financing, details of monetary and debt management policies, international capital flows and foreign exchange markets, and financial contagion.
Simulate the role of financial and monetary policies. Common macroeconomic models describe monetary policy as a stock M, and its time trajectory is independently selected by a central institution, without explicitly describing the business of implementing the policy. In fact, the change of money supply is due to the transaction between the government and the public, the conversion of goods or non-monetary financial assets into money, or similar transactions between banks and non-bank public. These transactions are the source of changes in the money stock. They are different and depend on how they change the wealth and securities of economic entities.
Valla's law and additional constraints. "Valla's Law" says that the sum of each excess demand function of a subject must be zero for each vector with each variable as an independent variable in any function. This "law" stipulates that the supply and demand tables of all markets notified by economic entities must meet the budget constraints. For example, the simulated asset market below refers to the value of each group that determines the demand for assets and wealth. The sum of the demand for several kinds of asset ownership by the family at the end of the period is equal to the demand for wealth by the family at the end of the period. This means that asset demand, for example, the sum of the partial derivatives of any interest rate must be equal to the partial derivatives of wealth demand for the same variable.
My collaborator William Brenard and I found (1968) that this consistency requirement is not always clearly observed in the theoretical and statistical models of financial markets. If there is no clear asset function for various assets, it may behave strangely in a way that the model builder did not expect. For example, if the demand for money is negatively related to interest rate, but not to the demand for total wealth, it means that the function of non-monetary assets carries the mirror image of the influence of interest on money. The best way is to write all the functions clearly, even if one of them is redundant, and put the same independent variable in all the functions.
(B) Summary and expectation of micro-foundation
J R Hicks's 1935 paper has always been the inspiration and challenge for me and many other monetary economists. It encourages us to find out the nature and function of money and promissory notes for future payment of money under the will of people to hold assets. By understanding these foundations, we can find the observable determinants of the demand for money and money substitutes. This exploration of the micro-foundation of monetary theory promotes the inventory theory model of transaction medium demand (baumol,1952; Tobin,1956; Miller and Earle, 1966) and portfolio selection model (Tobin, 1958). It's not over yet. I think the reason is that, in the basic paradigm of economic theory, it is difficult to explain why people want to keep paper that has no intrinsic contribution to utility or technology, and paper has positive value when it is exchanged with goods and services. I can't solve that profound problem, and I don't think it is a prerequisite for the actual monetary theory.
For this and other reasons, I admit that the macroeconomic model I advocate is only loosely related to the optimization behavior of each subject. According to an ancient tradition, the whole economic structure equation is a mixture of individual behaviors and the sum of many different individuals. This is a practical method, while the other two methods have serious shortcomings. One method is to keep the preferences of all subjects and the property diversity allowed in the complete equilibrium model; Optimizing the weak constraint on individual excess demand does not mean the constraint on urban land demand table at all. Another way is that all economic entities are similar to each other or divided into two or three categories (such as old people and young people), which are the same within one category, but different from each other in an artificial way. Although this kind of structure (such as Samuelson's 1958 intergenerational model) is promising and produces some instructive metaphors, they are still so abstract and arbitrary that they are useless for policy analysis and economic lecturer model construction.
In today's economics, another influential methodological trend is to attach importance to information and expectation, so it is necessary to establish such a model, in which the behavior of economic agents is based on information about the present and the future, and according to the model itself, economic agents obtain these information (Lucas, 1976). This is a good principle, but my own efforts to design and improve the structure have different purposes and priorities. I think the system shown in this lecture contains expectations. Like any behavior of financial returns, I think the attractiveness of various assets to depositors and portfolio managers depends on their estimation of the joint probability distribution of asset returns and capital gains. The consequences of these expected changes can be traced back, for example, as long as emotions, confidence and "anger" are changed from other places, Keynes believes that this is usually the case. Lucas' principle of reasonable expectation can be respected in dynamic application and simulation, but I haven't done it in the work I want to report here.
(c) Macroeconomics and overall equilibrium
Kenneth j. arrow's Nobel lecture in 1972 is a good exposition of the general equilibrium theory, which recognizes both its strength and its limitations. If all commodities have a completely clean market at the same time, including future and accidental markets, there will be no macroeconomic problems, no need for money, and fiscal and monetary stability policies are useless. Theorists who take full and complete equilibrium as their starting point will naturally interpret incompatibility as "market failure". Arrow discussed the problem that the decentralized competitive market can't provide collectives or commodities in the optimal quantity. In my opinion, the commercial nature of the common monetary accounting unit and the generally accepted medium of exchange is one of the reasons why all equilibrium paradigms bring trouble after accommodating money. But the deviation from that paradigm I want to emphasize is the deviation from the stage that led to macroeconomic theory and policy, and it is also the deviation emphasized by Keynes. The deviation is that there is no future market for any commodity except money itself, and of course there is no market for occasional payment. Keynes said (page 16936, page 2 10-2 12).
"Personal savings action means-for example-deciding not to eat dinner today. But it doesn't need to decide to buy a pair of boots for dinner a week or a year later, or to consume anything on any specified date. Therefore, it reduces the business of preparing dinner today and discourages the business of preparing for future consumption actions. This is not to replace the current consumption demand with the future consumption demand, but the net reduction of this demand ... If saving is not only to curb the current consumption, but also to give a specific order to the future consumption, the effect may indeed be different. Because in that case, the expectation of waiting for the investment to get some future income may increase, and the resources released for preparing for the present consumption may turn to preparing for the future consumption. ...
So the trouble is that instead of replacing the current consumption with a specific new consumption, the economic activities needed to prepare for the new consumption are as much as those needed for the current consumption with a value equal to savings, but a desire for "wealth" itself, that is to say, there is no time to consume the potential of goods without regulations.
In short, finance and capital markets are highly imperfect coordinators of savings and investment at best, and I doubt that reasonable expectations can make up for this defect. This kind of disharmony is a basic source of macroeconomic instability and macroeconomic stability policies. The current macroeconomic theory may pay too much attention to the labor market, and Keynes also found that competition could not coordinate the supply and demand of labor.
(4) Statistics of capital flow and stock
The national income account developed between the two wars provided data for testing and estimating the models of Keynes and later macroeconomists. Both theory and data are mainly aimed at traffic and its relationship. Capital flow accounts, especially those compiled by the Federal Reserve since 1949, provide useful stock and flow data for the theoretical model of financial markets. We try to understand and interpret these observations.
In tables 1 and 2, I have listed the data of 1979, which is less than the number of departments (columns) and asset categories (rows) actually reported by the Federal Reserve System. There are 9 departments and 1 assets in the table1,which is the total degree of a model that our team at Yale has been trying to estimate (Baxter et al., 1980). In Table 2, the data are further summarized into four departments and four assets, which is as close as possible to the theoretical model I want to discuss here.
In the format of these tables, one line represents a department's balance sheet (stock) or the source and use of funds (flow). A line allocates the stock or flow of an asset between the supply department and the demand department. The task of theory and estimation is to explain each column with some functions, which link the decisions of various departments on securities and savings with related variables, and interpret each row as a cascade of market clearing equations.