How exactly can recessions be predicted?


Philip Paul

To person

Philipp Paulus has been a research assistant at the Institute for Economic Policy at the University of Cologne since September 2002. Prior to that, he had worked for several years in the economic analysis department of an investment bank in London and Zurich since graduating from the University of Cologne in 1995.

Hardly any economic news comes without a reference to the economy. It seems to describe the state of the economy. But what do the numerous "economic barometers" from Ifo to ZEW really tell us?


    "It's a recession when your neighbor loses their job; it's a depression when you lose your own job." - Harry S. Truman, US President (1945-1953)
Hardly any economic news comes without a reference to the economy. In general, it seems to describe the state of the economy: If the economy is doing well, there are more jobs, higher salaries and general prosperity increases. If the economy is not going well, it is the other way around.

In economics, the economy is defined more narrowly: Here it shows the short-term state of economic output, i.e. the sum of all that is newly produced, while the long-term trend is captured by the concept of growth, which can be positive and negative. Economic data is subject to constant change. This is due to the nature of things: economy, business cycle and growth are the result of billions of individual human decisions worldwide. So it would be a pure coincidence if everything kept developing steadily. Therefore there is the long-term trend, also called the growth path, while the fluctuations denote the economy.

Scientists have identified different frequencies of cycles in the economy in previous economic development, but the business cycle describes two alternating phases, each lasting a few years, namely an upswing (= acceleration of the growth of annual economic output) and recession (= decrease in the growth of annual economic output) Economic performance, but still positive growth rates). In extreme cases, there is even a boom (very high growth rates in the economy) or a depression (the economy is shrinking, i.e. negative growth rates).

It is also important to distinguish between two components in economic and growth developments. On the one hand there are changes in real economic data and products (e.g. more cars are produced), on the other hand there are changes in their prices (the same number of cars costs more). If prices generally rise on average, there is inflation, in the rarer case of falling prices there is deflation. Most of the time, both variables - real production and prices - play a role in the economy and are mutually dependent. In the seventies there was even a phenomenon of "stagflation" i.e. there was less production, but at higher prices, which of course is a very bad economic situation.

As mentioned above, the economy is the result of many individual decisions. Economics tries to analyze the causes and effects of the business cycle and to discover laws. Sometimes neighboring sciences help: Sunspot phenomena, which astrophysicists can identify and calculate centuries back, can have caused bad harvest years and thus explain economic crises in times when the agricultural sector was more important than it is today. Observations from biology and psychology on the behavior of social groups can explain the herd instinct that can often be observed on the stock market, which can cause price collapses and (in more modern economies) even an economic crisis - up to and including a depression. The "Great Depression" in the USA from 1929-1933 has one of its causes here.

Two areas are of central interest in business cycle research. On the one hand, there is economic policy, which in some cases tries to keep economic fluctuations as low as possible - above all to mitigate negative trends and recessions. However, this is also risky, because it is precisely through this endeavor that it can in turn contribute to the fluctuations: unlike with millions of individual decisions, politicians can also move many economic variables at once through individual decisions, e.g. through subsidies or taxes. Even a central bank, which decides on monetary policy, controls an essential part of the economy through interest rates. The second central area is the forecasting of the future economic development, which can be useful for economic policy as well as for private purposes (e.g. investments).

Economic forecasts

    "There are three kinds of lies: lies, damn lies, and statistics." - Benjamin Disraeli, British Prime Minister in the 19th century, after Mark Twain
It would be very helpful in dealing with the economy to know where it is headed in the near future. For example, if we knew whether economic output continued to rise, we could bet on rising stock markets and buy stocks, politicians would consider higher tax revenues in budget planning, and so on. The problem is: an economic forecast, like any statement about the future, is uncertain. The problems start with the mere determination of where we are currently in the economy. Economic data must first be collected, processed and standardized, which means that we only have initial estimates of the economy when a few months of the year in question have passed. These estimates are revised again and again for past years, the more data are available.

This also applies to the economic forecasts of the large research institutes, central banks, but also private financial institutions and associations. In the best case, they can make statements about short-term trends; however, they come across exact numbers by chance. To determine the latest data and forecasts, the researchers use surveys conducted by companies and trade associations, as well as mathematical statistical methods. Survey results on moods and economic opinions are then numerically converted into economic indices (sometimes also "economic barometer"), for example by the Ifo Institute in Munich or the Center for European Economic Research (ZEW) in Mannheim, the monthly reports on the current economic situation and short-term trends publish.

Economic forecasts usually have a horizon of no more than two years. Anything beyond that are long-term forecasts of growth and general structural developments that are becoming increasingly uncertain. This category then includes, for example, statements on demographic development up to 2050.

Ultimately, economic forecasts are limited, but often quite helpful means of orientation in economic policy and economic research. Often it is enough to know whether the economy is expanding further or whether there is a threat of a recession so that appropriate measures can be taken, from economic policy to private investment decisions and corporate strategies. And, as is well known, stock exchanges always react not only to real events but also to opinions. Economic forecasts - even if they are wrong - can then, in a kind of "self-fulfilling prophecy", trigger exactly the economic effect that they expect.