Shaping earnings mobility: policy and institutional factors.

• Author: Sologon, Denisa Maria
• Position: Report

1. Introduction

   While much of the welfare literature focuses on cross-sectional inequality, the underlying dynamics or mobility of individuals within the distribution has significant implications in relation to welfare associated with being in different parts of the distribution. Earnings mobility represents a very important aspect for understanding earnings inequality.

   While earnings mobility can be welfare decreasing as a result of greater risk (Atkinson, Bourguignon, and Morrisson, 1992), it can also be seen as a mechanism for reducing long-term differentials, acting as a bridge between short-term and lifetime earnings inequality. For example, Friedman (1962) indicates that higher earnings mobility

   * is sign of a dynamic, more flexible and efficient economy,

   * contributes to equality of opportunity;

   * ceteris paribus, is expected to make the distribution of lifetime income more equal.

   The implication of the latter comment is that through "offsetting mobility", higher mobility can reduce inequality of income measured over a longer period of time, such as lifetime income or "permanent" income, despite a rise in annual inequality, resulting in a positive impact on long-term social welfare. However, it depends on the trade-off between long-term inequality and income variability or risk aversion (Creedy and Wilhelm, 2002; Gottschalk and Spolaore, 2002).

   This study focuses on the comparative analysis of earnings mobility in Europe. Of course, while wider income definitions such as disposable income are better determinants of welfare, employee earnings are one of the most important factors of disposable income and thus warren a distinct examination. Also, comparative studies, by comparing and contrasting the situation in different countries can facilitate a greater understanding in relation to the structure of mobility. Nevertheless, the literature in this area is relatively limited with only a few consistent comparative studies on earnings mobility, due to the lack of sufficiently comparable panel cross-country data (Aaberge, Bjorklund, Jantti, Palme, Pedersen, Smith, and Wannemo, 2002; Brukhauser and Poupore, 1997; Brukhauser, Holtz-Eakin, and Rhody, 1997; Fritzell, 1990; OECD, 1996, 1997; Hofer and Weber, 2002; Van Kerm, 2004).

   It is not, however, sufficient to describe the structure of earnings mobility. It is also instructive to understand the driving forces behind differences in earnings mobility across countries. Studies examining this are even scarcer. To our knowledge, the only example is Aaberge, Bjorklund, Jantti, Palme, Pedersen, Smith, and Wannemo (2002), comparing income inequality and income mobility in the Scandinavian countries and the U.S. during 1980-1990, to explore some of the factors driving mobility. They find that relative income changes are associated with changes in labour market and marital status in all four countries, but the U.S. record the largest magnitude of such changes. Thus there is a large gap in the literature analysing the factors shaping earnings mobility.

   Our paper attempts to fill part of this gap by exploring the role of labour market policy and institutional factors in understanding cross-national differences in earnings mobility across 14 EU countries. So far, at the EU level, no study explored the driving factors behind this labour market outcome in a comparative manner.

   In our study, mobility is regarded as the opposite of persistency, and can be interpreted as the opportunity for the poor to improve their relative income position in a lifetime perspective. We do not take the stand that mobility is necessarily good, but that the lack of it is bad, as it signals a lack of opportunity to move in the earnings distribution over the lifetime: in the absence of mobility the same individuals are stuck at the bottom of the distribution, hence annual earnings differentials are transformed into lifetime earnings differentials. Understanding the factors that enhance earnings mobility represents a step forward towards designing policies and institutions that enable low-wage workers to escape low-wage jobs and improve their position in the distribution of lifetime earnings.

   This question is highly relevant given the economic reality of the 1990s in Europe: the implementation of the single market (1992) and the preparation of the single currency (Maastricht criteria adopted in 1993) increased the pressure on the European labour markets to change. Starting in the early-1990s under the influence of the 1994 OECD Job Strategy, Europe has been moving towards more flexible labour markets through the implementation of employment-friendly reforms, expected to worsen the trade-off between a strong employment performance and a more equal distribution of earnings (OECD, 2004). The pace of change, however, was different across Europe (Palier, 2010), reinforcing the expectation of increased country-heterogeneity with respect to the labour market structure and the distribution of labour market income across Europe.

   Using the predicted earnings mobility from Sologon and O'Donoghue (2010a) and Sologon (2010) together with OECD data on institutional factors for 14 EU countries, we apply a nonlinear least squares method to explore the relationship between earnings mobility and labour market policy and institutional factors.

   Our study is divided into 5 sections. Section 2 describes the comparative pattern of earnings mobility across Europe. Section 3 discusses the link between institutions, labour market shocks and earnings mobility. Section 4 uses nonlinear least squares regressions to assess the linkage between these factors and earnings mobility, while section 5 concludes.

2. Earnings mobility

   There are a number of indicators that can be used to measure earnings mobility. Sologon and O'Donoghue (2010b) and Fields and Ok (1999) discuss the characteristics of a number of these measures, namely the Dickens, Shorrocks and Fields indicators. Rank measures best reflect movement in the earnings distribution between periods, traditionally derived from the transition matrix approach between income groups. This approach, however, does not capture the movement within each income group, thus potentially underestimating the degree of mobility. Dickens (2000a) computes the ranking of individuals in the wage distribution for each year and examines the degree of movement in percentile ranking between years. Shorrocks (1978) measures mobility as the relative reduction in the weighted average of single-year inequality when the accounting period is extended. In the literature it is usually classified among the measures of mobility as an equalizer of longer-term differentials. In recent years, however, this measure is criticised as it fails to capture the equalizing effect, failing to quantify the direction and the extent of the difference between inequality of longer-term income and inequality of base year income, treating equalizing and disequalizing changes in the same way (Benabou and Ok, 2001; Fields, 2008). Fields (2008) proposes an alternative index which captures mobility as an equalizer/disequalizer of longer-term incomes, circumventing the limitation of the Shorrocks index.

   In this study we wish to understand the driving factors associated with differential mobility across countries and over time. Most mobility indices, however, capture transitions between periods, which may make it difficult to link them with the yearly policy and institutional factors. One solution is to use a mobility index that reflects one particular period. Building upon the terminology introduced by Friedman and Kuznets (1954), individual earnings (i) in each year (t) are composed of a permanent and a transitory component, assumed independent of each other (Weizsacker, 1993):

   ln[(Earnings).sub.it] = [Permanent Earnings.sub.it] + [Transitory Earnings.sub.it] (1)

   The permanent component of earnings reflects personal characteristics, education, training and other systematic elements. The transitory component reflects individual random factors (e.g. illness and accident) and random changes in the market conditions in a particular period. It is expected to average out over time, with no influence on permanent earnings.

   Earnings mobility is closely related to these two components of earnings inequality. A large contribution of the permanent component implies that individual earnings are highly correlated over time and individuals do not change their income position to a large extent experiencing low rates of earnings mobility. Therefore, changes in earnings mobility are determined by the extent to which changes in cross-sectional inequality are driven by changes in permanent or transitory variance.

   The measure of mobility is derived from decomposing earnings inequality into its permanent and transitory components. Taking the variance of ln(earnings) in (1), under the independence assumption between permanent and transitory earnings, we obtain:

   Var(ln[{Earnings)).sub.c] = [Permanent Variance.sub.c] + [Transitory Variance.sub.t] [Earnings Inequality.sub.t] = [Permanent Inequality.sub.t] + [Transitory Inequality.sub.t]

   Earnings Immobility Ratio = Permanent Variance/Transitory Variance (3)

   The degree of immobility measured by the ratio between permanent and transitory inequality (3), in the tradition of Kalwij and Alessie (2003), is the measure of mobility that we use in this study. This measure offers a summary of the evolution in the structure of inequality: an increase in the immobility ratio indicates a decrease in earnings mobility, equivalent with an increase in the relative share of permanent differentials in the overall inequality. This mobility measure captures non-directional earnings movements and can be interpreted as the opportunity to improve one's position in the distribution of lifetime earnings.

   Data on earnings mobility using these measures is taken from Sologon and O'Donoghue (2010a, 2010b) and Sologon...