The Time Lag in Annual Household‐based Income Measures: Assessing and Correcting the Bias

Abstract

Annual income data are typically provided with a time lag. This article reviews several ways of dealing with this time lag in the construction of annual household-based income measures for individual economic well-being. It also proposes an alternative method that yields better estimates for equivalized household income, especially in the case of household composition change. Next, the two most commonly applied income measures are compared to this alternative measure with empirical income data from the European Community Household Panel. This comparison reveals that ignoring the time lag and household changes leads to substantial bias in income and poverty estimates and to erroneous conclusions about the determinants of poverty entry. The evidence in this article will be useful to researchers who want to make a well-informed choice between different annual income measures. Copyright 2008 The Authors.

Measuring the effect of spell recurrence on poverty dynamics—evidence from Spain

“…11 Results on the conditional probability of transition are reported in Table 2. The first row of these conditional probabilities indicates the individual probability of remaining in poverty in two consecutive interviews i.e.…”

“…Thus, it includes 19,129 individuals of which 15,096 (79%) are adults and 4,033 (21%) are children below 16 years of age (see Table 11 in the Appendix). 8 For the purposes of our research, we use the standard European or EU definition of poverty, thus an individual is poor if total household income of the household she 6 See Debels and Vandecasteele [11] for a discussion of the empirical relevance of ignoring this time lag in analysing poverty dynamics in the European Union. 7 We eliminate between 1 and 2% of individuals due to the lack of complete interview-see Table 10 in the Appendix.…”

Public cash transfers and poverty dynamics in Europe

“…In addition, the 11 Descriptive statistics for the country-level variables are presented in Appendix Tables 7 and 8. 12 For this purpose, we use the modified OECD equivalence scale which assigns the value of 1 to the first adult in the household, the value of 0.5 to each additional person aged 14 or over who is living in the household, and the value of 0.3 to each child aged less than 14 (Eurostat 2003b). 13 A detailed discussion of the time lag in annual household-based income measures and possible solutions to solve this problem can be found in Debels and Vandecasteele (2008). The receipt of public transfers shows the shares of the poor and non-poor who receive them.…”

Measuring Income in Household Panel Surveys for Germany: A Comparison of EU-SILC and SOEP

“…Alongside these indicators based on cross-sectional data measuring income inequality at a single point in time or in a time series, the panel data collected in EU-SILC and other similar 13 22 While this is the standard approach in income distribution analysis, it may lead to biased measures of equivalent income due to the different points in time referred to with income (previous year) and household composition (current interview month) (see Debels & Vandecasteele 2008). This effect should not, however, significantly affect the present study, whose aim is to analyze the comparability of distribution and mobility analyses based on EU-SILC and SOEP, since both surveys are equally affected.…”