24 February, 2016

Methods

The main features of the chronology of the Spanish Business cycle dating committee are the following: 

  • The chronology determines dates of peaks and troughs in economic activity in Spain.
  • Recessions start at the peak and end at the following trough. Expansions start immediately after the trough and end immediately before the following peak.
  • Within each cyclical phase, recession or expansion, there may be transitory fluctuations not considered to be fully fledged cyclical shifts. For example, within a given expansion (recession) there can be a brief period of economic decline (improvement).
  • The CF (for dating committee, Comité de Fechado in Spanish) does not use a mechanical rule to determine the minimum length of each cyclical phase.
  • Recessions are periods of decline widely spread across the economy. This definition does not rule out that some parts of the economy may be experiencing economic growth.
  • The CF uses data on several economic indicators to date the business cycle. Fluctuations in Gross Domestic Product or GDP (PIB in Spanish) are particularly important, but are not the only source of information. The CF also takes into account labor market developments, industrial production, and other indicators described in more detail below. We track multiple indicators to resolve timing conflicts between alternative indicators and to improve the precision of the dating.
  • The CF usually waits several months before announcing cyclical shifts to avoid having to revise the chronology a posteriori. There are two main reasons for this delay. First, initial data releases are usually revised by statistical agencies, often by a considerable amount. Second, several months worth of data is necessary to assess whether there has really been a phase shift or whether it is simply a transitory fluctuation within a given cyclical phase.
  • In order to mitigate the delay in dating cyclical shifts and to provide real-time information about the current state of the business cycle, the CF provides an indicator of economic activity at a monthly frequency. This tool (and others we expect to provide in the future) is provided separately from the chronology. They should not be taken as determining the dates of peaks and troughs per se.

List of indicators frequently used to construct the Spanish business cycle chronology:

  • Real Gross Domestic Product: GDP at market prices, chain index of volume, reference 2010. Series adjusted for seasonality and calendar effects. Source: Quarterly National Accounts (QNA) Base 2010, Spanish National Statistical Office (INE). Sample period: from 1995Q1.Backcasted until 1970Q1 using quarterly growth rates of GDP, volume index, reference 2000, QNA Base 2000, published by INE.
  • Real Private Consumption: Final expenditure of households at market prices, chain index of volume, reference 2010. Series adjusted for seasonality and calendar effects, QNA Base 2010, INE. Sample period: from 1995Q1. Backcasted until 1980Q1 using annual growth rates of the corresponding variable of QNA Base 1995, and from 1970Q1 to 1979Q4 using annual growth rates of the corresponding variable of QNA Base 1986, published by INE.
  • Real Gross Fixed Capital Formation: chain index of volume, reference 2010. Series adjusted for seasonality and calendar effects, QNA Base 2010, INE. Sample period: same as real Private Consumption.
  • Real Export and Imports of goods and services: chain indices of volume, reference 2010. Series adjusted for seasonality and calendar effects, QNA Base 2010, INE. Sample period: same as real Private Consumption.
  • Real Gross Value Added by branch of activity: chain indices of volume, reference 2010. Branches of activity: Agriculture, Industry (includes energy), Construction and Services (both market and non market). Series adjusted for seasonality and calendar effects, QNA Base 2010, INE. Sample period: same as real Private Consumption.
  • Wage compensation per employee: Total wage compensation divided by the number of employees in full time equivalent jobs, whole economy. Thousands of current euros. Series adjusted for seasonality and calendar effects, QNA Base 2010, INE. Sample period: from 1995Q1. Backcasted until 1980Q1 using annual growth rates of the corresponding variable of QNA Base 1995, published by INE.
  • Employment: Full time equivalent jobs, in thousands of people. Whole economy and main branches of activity: Agriculture, Industry (includes energy), Construction and Services (both market and non market). Series adjusted for seasonality and calendar effects, QNA Base 2010, INE. Sample period: from 1995Q1.Backcasted until 1980Q1 using annual growth rates of the corresponding variable of QNA Base 1995, published by INE.
  • Unemployed: thousands of unemployed, total. Seasonally adjusted series. Methodology EPA-2005, population base (Censo) 2011. Source: Encuesta de PoblaciónActiva (EPA), INE. Sample period: Data in levels and unadjusted available since 2002Q1 using Censo 2011. We have applied to the last data point (2014Q3) the series of seasonally adjusted quarterly rates of change since 1990Q1 published by INE. Finally, the series have been further backcasted until 1976Q3 using inter annual rates of change based on methodology EPA-2005 and Censo 2001, published by INE.
  • Employment EPA: thousands of employed, total. Seasonally adjusted series. Methodology EPA-2005, population base (Censo) 2011. Source: Encuesta de PoblaciónActiva (EPA), INE. Sample period: same as Unemployed.