Do Firms Provide Wage Insurance against Shocks?

• Published date: 01 January 2016
• DOI: http://doi.org/10.1111/sjoe.12128
• Date: 01 January 2016

Scand. J. of Economics 118(1), 105–128, 2016

DOI: 10.1111/sjoe.12128

Do Firms Provide Wage Insurance

against Shocks?∗

G´

abor K´

atay

Banque de France, FR-75001 Paris, France

gabor.katay@banque-france.fr

Abstract

Implicit contract models imply that it is Pareto optimal for risk-neutral firms to provide

insurance to risk-averse workers against shocks. Using a matched employer–employee dataset,

I evaluate wage responses to both permanent and transitory shocks in Hungary, and compare

my results to similar studies on Italian, Portuguese, German, and French datasets. I find

that the magnitude of the wage response differs depending on the nature of the shock.

Broadly speaking, the wage response to permanent shocks is twice the size of the response

to transitory shocks. Unlike previous findings, my results show that full insurance against

transitory shocks is rejected.

Keywords: Matched employer–employee data; optimal wage contract; product market uncer-

tainty; risk sharing; wage insurance

JEL classification:C33; D21; J33; J41

I. Introduction

There is a consensus nowadays that the market for labour cannot be repre-

sented satisfactorily by a standard competitive model. In particular, workers’

wages are positively correlated with an employer’s ability to pay. However,

firms’ prof its (or value added) usually fluctuate to such an extent that if

wages were to follow them on a one-to-one basis, workers would experi-

ence excessively large variations in their compensation. On the one hand,

employees are generally viewed as risk-averse and, as such, they derive

negative utility from an increase in wage income variance. On the other

hand, owners of firms (shareholders) usually have better access to financial

markets, and can thus protect themselves more easily against idiosyncratic

risks by diversifying their portfolio of assets. Therefore, it follows that in

the presence of uncertainty about future market conditions, it is Pareto op-

timal for risk-neutral firms to provide risk-averse workers with insurance

∗Part of this work was carried out while I was visiting the European Central Bank, whose

hospitality is gratefully acknowledged. I am especially indebted to Frank Smets and Gabriel

Fagan for their useful comments and support. I would also like to acknowledge the critical

and extensive comments from Fabiano Schivardi and from the participants of the Wage

Dynamics Network.

CThe editors of The Scandinavian Journal of Economics 2015.

106 Do firms provide wage insurance against shocks?

against shocks (i.e., for managers to commit to pay a pre-agreed wage,

independently of product market fluctuations).

This paper follows the line of research that attempts to capture the

existence of, and to quantify the extent of, implicit wage insurance. The

first few studies on the subject used aggregate industry data (e.g., Gamber,

1988), while later, an increasing number of papers used individual data on

workers, but still defined shocks at the industry level (see, for example,

Beaudry and DiNardo, 1991; Weinberg, 2001; Devereux, 2005).

More recently, a number of papers have made use of matched employer–

employee micro datasets to quantify the impact of firm-level value-added

shocks on individual wages.1The present paper is most closely related to

Guiso et al. (2005), which uses Italian employer–employee data to show

that Italian firms provide full insurance against temporary fir m-level value-

added shocks, and only partial insurance against permanent shocks. By

replicating the empirical identification strategy of Guiso et al., Cardoso

and Portela (2009) and Le Barbanchon and Tarasonis (2011) report sim-

ilar results for both Portuguese and French data, respectively. In the case

of Germany, a recent study by Guertzgen (2014) shows that the average

German worker is fully insured even against permanent shocks.

In line with these studies, this paper presents an empirical analysis of

wage responses to firm-level shocks in Hungary, using a similar methodol-

ogy to Guiso et al. (2005). The advantage of re-examining wage sensitivity

to idiosyncratic shocks for Hungarian data is that the economic and in-

stitutional background in Hungary differs substantially from the Italian,

Portuguese, French, or German environments in many respects. Although

collective bargaining coverage has been on the decline in all of these coun-

tries, the time periods covered by these previous analyses were characterized

by a broadly regulated system of wage bargaining, with widespread mech-

anisms of contract extension and, consequently, high rates of trade union

coverage. According to OECD (2012), in 1990, trade union coverage in

these countries varied between 72 percent in Germany and 92 percent in

France. Even these days, sectoral (and, in Germany, regional) level wage

bargaining is still most prevalent, company-level agreements generally cover

only a limited proportion of employees (see, e.g. Du Caju et al., 2008).2

1Sector-specific or even country-level shocks are diversifiable by adopting a multi-sectoral or

international portfolio. Hence, it might be optimal for owners to insulate risk-averse workers’

wages from aggregate shocks as well. Moreover, even with purely aggregate shocks, there

might still be scope for risk sharing, provided that the owners are less risk-averse than

workers. The distinction between aggregate and idiosyncratic shocks is nonetheless relevant

as the amount of insurance that firms provide might be different for each type of shock.

2Despite these institutional features, the authors of these papers argue that the wage compo-

nent determined at firm level is significant enough for f irms to influence the wages of their

CThe editors of The Scandinavian Journal of Economics 2015.