This chapter is concerned mainly with the history of the econometric approach to business cycles. However, because the historical development of the approach is obviously associated with that of both business cycle theory and econometrics, the discussion must be broad enough to include the following issues, which could otherwise be dealt with separately.

The first issue is a fundamental methodological problem of econometrics. Given the fact that the data alone do not enable us to discover or reveal the laws governing them, it is obvious that a priori theory should play a role in the econometrician's search for truth. The extent to which economic theory should be introduced was a controversial issue in the earlier history of econometrics. Theory without statistical inference, empirical investigations without theory, and econometrics as a bridge between theory and observations are positions that could be roughly identified with such names as Keynes, Mitchell, and Koopmans, respectively. Even today, these positions can be found in various schools of econometrics. Later in the chapter, this issue will be discussed in the context of criticisms of the econometric approach to business cycles.

The second issue is related to the question of why and how the business cycle faded away as an independent subject of study. Starting in the late 1940s, the popularity of cycle theories began to decline, and they were eventually replaced by Keynesian macroeconomics. The usual explanation for this redirection of aggregate economics stresses both the advent of Keynesianism and historical experiences – for example, the fact that the postwar business cycle became mild and substantially stable (e.g., Bronfenbrenner, 1969).