Labour market flows: Accounting for the public sector
Introduction
In most European economies, around 20 percent of all workers are employed by the government. Government hire workers to produce goods and services. However, governments face different constraints than private-sector firms and are not driven by profit maximization. Hence, government employment and wage policies are driven by other objectives including: attaining budgetary targets (Gyourko, Tracy, 1989, Poterba, Rueben, 1998); implementing macroeconomic stabilization policy (Holm-Hadulla, Kamath, Lamo, Pérez, Schuknecht, 2010, Keynes, 1936, Lamo, Pérez, Schuknecht, 2013); redistributing resources (Alesina, Baqir, Easterly, 2000, Alesina, Danninger, Rostagno, 1999, Wilson, 1982); or satisfying interest groups for electoral gains (Borjas, 1984, Gelb, Knight, Sabot, 1991, Matschke, 2003). As a consequence, public-sector labour markets might behave differently from their private-sector counterparts.
The objective of this paper is to establish a number of key facts about the French, Spanish, UK and US labour market flows, focussing on the role played by the public sector. We do so by examining data from the French, Spanish and UK Labour Force Surveys and the US Current Population Survey (CPS) over the past 15 years. We chose these four countries because they are large countries with sizable public sectors, and have been recently facing pressure to reform their public sectors. Furthermore, because they have different labour market institutions, public-sector hiring procedures and wage policies and various weights on different industries, facts that are found to be common across the four countries should be seen as intrinsic characteristics of the public sector. While we do not attempt to explain these facts, we believe that they are an important first step to foster theoretical research on the topic. They can help economists understand the characteristics of the public sector and its policies, as well as provide a guideline of the empirical features that models with a public sector should reproduce and help in the calibration or identification of key parameters. We show that public-sector labour markets do indeed behave differently than the private sector. The size of transition rates into and out of public-sector employment are different and its cyclical pattern as well. Furthermore, the government hires mostly women, college graduates and older workers, which creates asymmetric exposures to public-sector policies for different workers.
In the last decade in European countries, public-sector employment was a key policy variable. Following budgetary constraints, many countries imposed measures such as hiring freezes layoffs of public-sector workers, as well as wage cuts or freezes that affected the retention of these workers (Glassner and Watt, 2010). Given the policy role that public-sector employment played during the last decade, a new wave of research constructs search and matching models of unemployment to study the labour market effects of public-sector employment and wages. Examples include Hörner et al. (2007), Quadrini and Trigari (2007), Afonso and Gomes (2014), Gomes, 2015b, Gomes, 2018, Michaillat (2014), Burdett (2012), Bradley et al. (2017), Albrecht et al. (2019), Bermperoglou et al. (2017) and Boeing-Reicher and Caponi (2017). Lying at the heart of these state-of-the-art models are the worker gross flows between private- and public-sector employment and non-employment.
However, the extensive literature that estimates and analyses worker gross flows has systematic ignored the role of the public sector. This literature has focused mainly on disentangling the relative importance of job-finding and job-separation rates in driving the unemployment rate. The most cited papers on the topic – Blanchard and Diamond (1990), Shimer (2012), Elsby et al. (2009) and Fujita and Ramey (2009) – study the US labour market, proposing different decompositions or examining the role of the time-aggregation bias. Also for the US, Borowczyk-Martins and Lalé (2019) distinguish between full-time and part-time employment, while Elsby et al. (2015) study the role of the participation margin. Smith (2011) proposes an out-of-steady-state decomposition and analyses the UK labour market. Gomes (2012) further analyses the UK labour market along other dimensions, such as education or labour force attachment, while Fujita (2010) concentrates on on-the-job search and job-to-job transitions and Carrillo-Tudela et al. (2016) on the extent of worker reallocation across occupations and industries and their cyclicality. In two comparisons of the UK and the US, Razzu and Singleton (2016) study the fluctuations of unemployment among men and women, while Gomes (2015a) examines the role of conditional transition probabilities and how they depend on the frequency of the surveys. Other papers focussing on the UK include Elsby et al. (2011) and Elsby and Smith (2010).
Several studies examine other European labour markets. Petrongolo and Pissarides (2008) compare the relative importance of job-finding and job-separation rates across France, the United Kingdom, Spain and the United States. Silva and Vázquez-Grenno (2013) focus on the role of flows in and out of permanent and temporary employment in Spain. Baussola and Mussida (2014) study Italian gross flows, concentrating on unemployment gender gaps. Charlot et al. (2018) split between employment in abstract, routine and manual occupations in France and the US. Other works examining the French labour market include Hairault et al. (2015) and Fontaine (2016). Hertweck and Sigrist (2015) study the German labour market and Daouli et al. (2015) the Greek labour market during the crisis. Despite looking at worker flows from different angles, all the papers in this exhaustive list have ignored the duality between the private and the public sectors.
In Section 3, we provide evidence on the size and cyclicality of the flows between public and private employment, unemployment and inactivity. France and the UK have larger public sectors than either Spain or the US. Over the last business cycle, public-sector employment was pro-cyclical in France, countercyclical in the US, and acyclical in Spain and the UK.
In Section 4, we quantify how government hiring and separations have contributed to unemployment fluctuations. We show that ignoring these flows in unemployment decompositions can potentially bias the relative importance of job-finding and job-separation rates, although in our sample, this bias turned out to be small. We find a relative split of 80-20 percent of the contribution of private- and public-sector employment to fluctuations in the unemployment rate in UK, 85-15 in France and of 90-10 percent in Spain and the US. We performed a counterfactual analysis and show that since 2008, if governments had kept the same hirings and separations from the previous years, unemployment rate would have been lower, by up to 1 percentage point, in the France and the UK, but it would have been higher in the US and Spain. In our view, this finding reflects the different macroeconomic policies conducted by governments in response to the Great Recession, with a larger focus on austerity policies by some European countries.
We document that jobs are safer in the public sector – aggregate job-separation rates are lower. In Section 5, we further investigate this result by using a multinomial logit model to estimate the differences in transition rates from employment to unemployment and inactivity from the two sectors, conditional on observable characteristics. The argument that public-sector jobs are safer is often used in policy discussions surrounding public-sector pay. However, while there are several papers estimating the wage differentials across sectors, there are no estimates of the value of the job-security.1 We use a simple back-of-the-envelope calculation to find the percentage of their wage that private-sector workers would be willing to forgo to have the same job-separation probability as in the public sector. In our preferred scenario, risk-neutral workers would pay 0.5 to 1.6 percent of their wage for the same job security, which can be seen as a lower bound for the insurance value of public-sector employment. Risk-averse workers without any savings mechanism would pay 1.0 to 2.9 percent of their wage, which can be seen as an upper bound. The value of job safety in the public sector is equivalent to between 0.4 to 0.7 percent of total government spending in France and between 0.2 to 0.4 percent in the UK, Spain and the US.
Section snippets
Labour market dynamics
In order to analyse labour market dynamics, we use some fundamental equations that describe the evolution of the stock of the employed in the private and public sectors (P and G) and the stock of the unemployed U. The pool of the inactive is denoted by I. Adding the four pools gives us the working-age population W, while the sum of employment and unemployment corresponds to the labour force L. The unemployment rate is defined as and the participation rate as .
Changes in private and
Average gross flows
Fig. 1 summarizes the average quarterly (monthly) worker flows over the 2003–2018 period for the three European countries (United States). It reports the stocks of workers in thousands (t) and as a percentage of the working-age population (p), as well as the number of people that change status every quarter (month) as a percentage of the working-age population (p) and as a transition probability or hazard rate (h). We restrict our analysis to the working-age population (16 to 64 years old). The
Why does the public sector matter?
To understand the effects of ignoring the public sector when decomposing unemployment fluctuations, consider the following example of an economy with a public sector that has extremely low turnover. By this, we mean a separation rate λGU very close to zero, as well as the hiring rate λUG. There are also no movements between public and private sector. This scenario translates into a public sector with fixed size unresponsive to the economic cycle. If one were to do a standard two-state
How safe are public-sector jobs?
The argument that public-sector jobs are safer is often used in policy discussions over public-sector wages. According to Gomes (2015b), the optimal design of the public-sector wage schedule should take job security into account. Safer jobs raise a job’s expected duration of a job and reduce the expected time spent in unemployment. Thus, the government should offer lower wages in order to keep the value of a public-sector job in line with that of the private-sector job. Hence, the estimation of
Conclusion
The objective of this paper was to establish a number of key facts about public- and private-sector labour market flows. It provides a picture of a wide range of information about worker gross flows from different angles, improving our understanding of the workings of these two labour markets. The main findings of this paper can be summarised as follows:
- •
In France and the UK, the public sector represents 21 and 23 percent of total employment, respectively. Spain and the US have smaller public
Acknowlgedgments
We would like to thank the editor and two anonymous referees. We would also like to thank Mike Elsby, José Ignacio Silva and Thepthida Sopraseuth for initial discussions and participants at the University of Girona seminar, NIERS seminar, Royal Economic Society Conference, Symposium of the Spanish Economic Association and the T2M Conference. Part of this research has been done when Idriss Fontaine was a Post-Doc at the department of Economics (THEMA) of the University of Cergy-Pontoise.
References (56)
- et al.
Interactions between private and public wages
J. Macroecon.
(2014) - et al.
Redistributive public employment
J Urban Econ.
(2000) - et al.
Public-Sector Employment, Wages and Education Decisions
(2019) - et al.
Exploring the public-private sector wage gap in european countries
IZA J. Eur. Labor Stud.
(2013) - et al.
The Public Sector Pay Gap in a Selection of Euro Area Countries
(2011) - et al.
The ins and outs of German unemployment: a transatlantic perspective
Oxf. Econ. Pap.
(2015) A theory of countercyclical government multiplier
Am. Econ. J.
(2014)The ins and outs of UK unemployment
Econ. J.
(2011)- et al.
Public-sector employment in an equilibrium search and matching model
Econ. J.
(2019) - et al.
Redistribution Through Public Employment: The Case of Italy
(1999)