«Discussion Papers represent the authors‘ personal opinions and do not necessarily reflect the views of the Deutsche Bundesbank or its staff. ...»
unemployment and family firms
(European University Institute)
Discussion Papers represent the authors‘ personal opinions and do not
necessarily reflect the views of the Deutsche Bundesbank or its staff.
Editorial Board: Daniel Foos
Deutsche Bundesbank, Wilhelm-Epstein-Straße 14, 60431 Frankfurt am Main,
Postfach 10 06 02, 60006 Frankfurt am Main Tel +49 69 9566-0 Please address all orders in writing to: Deutsche Bundesbank, Press and Public Relations Division, at the above address or via fax +49 69 9566-3077 Internet http://www.bundesbank.de Reproduction permitted only if source is stated.
ISBN 978–3–95729–132–5 (Printversion) ISBN 978–3–95729–133–2 (Internetversion) Non-technical summary Research Question In this paper, I calculate unemployment multipliers of expenditure and revenueside ﬁscal consolidation policies in the presence of sectoral heterogeneity, ie. in the presence of family and non-family ﬁrms.
Many countries have experienced a rise in government debt since the recent crisis, which increases the need for ﬁscal consolidation, while, unemployment has simultaneously reaching historically high levels. Fiscal consolidation also aﬀects unemployment. Our knowledge of unemployment ﬁscal multipliers is not complete, however.
Concerning expenditure-side unemployment multipliers, evidence about both the sign and the magnitude is mixed. Moreover, there are not yet any revenue-based unemployment ﬁscal multipliers.
I develop a model, based on a standard New Keynesian framework with search and matching frictions, as an innovation, with sectoral heterogeneity, ie. family and non-family ﬁrms. The motivation for this is that family-owned ﬁrms employ a signiﬁcant share of the labor force, and these ﬁrms behave diﬀerently in the labor market. For the purpose of illustration, the model is calibrated to match data of those European countries where family ﬁrms are present more than average in the labor market.
Contribution Regarding the existing literature, my results are consistent with the suggestion that a tightening of ﬁscal policy on the expenditure side increases unemployment.
Moreover, consistent with Ball et al. (2013), a cut in government consumption implies a more pronounced eﬀect on unemployment than tax-based adjustments (only at peak though). But the relevant authors consider a narrative approach, which means they are unable to compare tax policies, while my paper is the ﬁrst to provide estimates of unemployment multipliers of diﬀerent tax policies. Furthermore, sectoral heterogeneity on the ﬁrm side might explain the gap between theoretical and empirical multipliers reported, but unexplained by Monacelli et al. (2010).
Results My model predicts that ﬁscal austerity raises unemployment. At peak, the largest increase in unemployment is implied by a cut in government consumption. Nevertheless, a hike in employees’ labor income tax, cumulatively, implies the same size of increase in unemployment as the government consumption cut does. A higher employer social security contribution is, however, less costly in terms of employment than an increase of the same scale in employees’ labor income tax hike. Both at peak and cumulatively, unemployment reacts least when the budget is consolidated by increasing the rate of value added tax. However, there are trade-oﬀs a policymaker must face, as the increase in value added tax results in the largest decline in consumption.
Sectoral heterogeneity seems to play a crucial role, unemployment multipliers are very diﬀerent with and without it. Multipliers of labor income tax policies and government consumption multipliers are usually biased downwards, while the consumption tax multipliers are often biased upwards, when homogeneous ﬁrms are considered. Thus, ignoring sectoral heterogeneity might lead to incorrect policy conclusions, although, according to my results, budget consolidation is always least harmful for employment when it is implemented by increasing consumption tax revenue.
Nichttechnische Zusammenfassung Forschungsfrage In der vorliegenden Arbeit werden Multiplikatoren der Arbeitslosigkeit bei ausgabenund einnahmenseitigen ﬁskalischen Sparmaßnahmen berechnet. Dabei wird eine sektorale Heterogenit¨t, d. h. das Vorhandensein familien- und nicht-familiengef¨hrter a u Unternehmen, zugrunde gelegt.
I calculate unemployment multipliers of ﬁscal consolidation policies in a standard, closed-economy New Keynesian framework with search and matching frictions, and, as an innovation, in the presence of sectoral heterogeneity. Family and non-family ﬁrms behave diﬀerently in the labor market and are diﬀerently managed. This latter assumption is modeled by the inclusion of intangible capital in the family sector. The model is calibrated to match European data on countries with a large percentage of family ﬁrms in the labor force. I ﬁnd that ﬁscal austerity raises unemployment. Both at peak and cumulatively, unemployment reacts least when the budget is consolidated by increasing the rate of value-added tax. At peak, the highest increase in unemployment is induced by a cut in government consumption, but, cumulatively, a hike in employees’ labor income tax is just as costly in terms of employment. There are trade-oﬀs, however, which a policymaker must face, as the value-added tax increase results in the steepest decline in consumption. Sectoral heterogeneity is crucial; multipliers of labor income tax policies and government consumption multipliers are usually biased downwards, while the consumption-tax multipliers are often biased upwards. Thus, ignoring sectoral heterogeneity might lead to incorrect policy conclusions.
Keywords: ﬁscal austerity, government consumption, labor income tax, consumption tax, social security contribution, unemployment multiplier, sectoral heterogeneity, family ﬁrms, intangible capital JEL codes: E22, E24, E62, J64 Acknowledgement: The project was partly completed during my internship at the Deutsche Bundesbank between January and March 2014. I would like to thank Angela Abbate, Daniel Baksa, Tamas Brieglevics, Fabio Canova, Nir Klein, Martin Kliem, Stephane Moyen, Evi Pappa and Nikolai Staehler for their valuable comments and support. I am also grateful for the comments made by participants of presentations held at the European University Institute and the Deutsche Bundesbank, also from persons, who attended the Rimini Conference in Economics and Finance (June 2014), the Nyari Muhely of the Hungarian Academy of Sciences (August 2014) and the Debunking Austerity conference organised by the Scuola Superiore Sant Anna (November 2014). Any remaining errors are my own.
A technical appendix containing all the steady state and loglinearized equations is available on my website: https://sites.google.com/site/munkacsizsu/research.
Doctoral Researcher, European University Institute, e-mail: Zsuzsa.Munkacsi@EUI.eu 1 Introduction Since the onset of the recent crisis, many countries have experienced a rise in government debt, thus making urgent ﬁscal consolidation. At the same time, unemployment in many countries has risen to historic heights. We know that ﬁscal consolidation further aﬀects unemployment, but we do not know the precise mechanism behind this phenomenon. This is the question which I address in this paper.
In particular, I calculate unemployment ﬁscal multipliers that show the eﬀects of temporary ﬁscal austerity policies on the unemployment rate. My paper contributes to the literature in two ways (Subsection 1.1). First, there is a debate on the sign and the magnitude of expenditure-side unemployment ﬁscal multipliers. Second, as yet there are no revenue-side unemployment ﬁscal multipliers. The intent is to ﬁll both of these gaps of the literature.
The multipliers are based on a standard closed-economy New-Keynesian model with search and matching frictions, and as a innovation, with family and non-family ﬁrms on the ﬁrm side. The motivation behind this sectoral heterogeneity is that ﬁrms are not homogeneous, as the standard macroeconomic models often assume.
Empirical ﬁrm-level evidence from the corporate ﬁnance literature indicates that family ﬁrms behave diﬀerently in the labor market than non-family ﬁrms do (Subsection 1.2).
1.1 Unemployment ﬁscal multiplier literature Unemployment ﬁscal multipliers show changes in the unemployment rate after a temporary ﬁscal policy shock. The literature focuses on the expenditure side of the budget. Monacelli et al. (2010) ﬁnd that an increase of 1 percent of GDP government consumption decreases unemployment by 0.6 percentage points at peak3, but their theoretical multiplier is much lower, around 0.2 percentage points. Similarly, Edelberg et al. (1999), Fatas-Mihov (2001), Gali et al. (2007), Forni-Gambetti (2010) and Mayer et al. (2010) claim that loosening ﬁscal policy implies an increase in hours or employment.4 In contrast, Bruckner-Pappa (2012) suggest that an increase in government consumption raises unemployment.5 Their model, which includes price rigidity, labor force participation, as well as short- and long-term unemployment reproduces their empirical ﬁndings. Gomes (2009) claims that a shock to government consumption - without the inclusion of government employment in the model6 - implies a close-to-zero eﬀect on unemployment. Moreover, Dallari (2014) ﬁnds that the impact multipliers of a cut in government consumption might be positive or negative.7 Based on a VAR of the US.
Edelberg et al. (1999), Fatas-Mihov (2001) and Gali et al. (2007) consider 1-leisure as equal to hours or employment in their framework, without, however, explicitly modeling unemployment, while Forni-Gambetti (2010) provide empirical evidence based on a structural, large dimensional, dynamic factor model.
A 10 percent increase in government consumption results in an increase of 0.2-0.5 percentage points in unemployment at peak by estimating structural VARs of several OECD countries.
A DSGE model with search and matching frictions, public employment and wages.
Multipliers vary from −4.5 to 8.7, using a panel structural VAR of European countries.
Thus, there is a debate on both the magnitude and the sign of the unemployment multiplier of government consumption.
Most papers concentrate solely on government consumption, although diﬀerent government expenditure items might imply diﬀerent employment eﬀects.8 Bermperoglou et al. (2013) ﬁnd that a decrease of 1 percent of GDP in government consumption, government investment and public vacancies increase unemployment by 0.8, 0.8 and 1.8 percentage points respectively, while a decrease of 1 percent of GDP in public wages reduces unemployment by 0.1 percentage points.9 A New Keynesian model with labor force participation, short- and long-term unemployment and public employment provides similar theoretical responses. At the same time, Gomes (2009) shows that an increase in public vacancies results in higher unemployment, while an increase in public wages declines unemployment. Moreover, Pappa (2009) ﬁnds that a positive shock to public employment decreases total employment in several US federal states.10 Also, Dallari (2014) demonstrates that after cuts in government investment, unemployment multipliers of diﬀerent countries have diﬀerent signs and magnitudes.11 The aforementioned studies rely on aggregate government expenditure data. At the same time, the narrative approach achieves identiﬁcation using historical records of the policy decision-making process. Both Guajardo et al. (2011) and Ball et al.
(2013) suggest that ﬁscal tightening increases unemployment.12 Hernandez de CosMoral-Benito (2011) report similar results.13 While there is no consensus in the literature regarding the sign or the magnitude of expenditure-side unemployment ﬁscal multipliers, I am not aware of any unemployment multipliers of tax policies.14 Although Ball et al. (2013) claim Some papers speciﬁcally focus on military spending, which is not discussed here, see e.g.
Rotemberg-Woodford (1992), Ramey-Shapiro (1998), Burnside et al. (2004) and Ramey (2009).
Based on a SVAR with sign restrictions of the US.
Using a structural VAR with sign restrictions on US aggregate and state level data.
On impact −1.7-4.1 in Europe.
Both papers use a sample of OECD countries. Guajardo et al. (2011) claim that, two years after the shock, a ﬁscal consolidation of 1 percent of GDP implies a 0.3 percentage point increase in unemployment, while Ball et al. (2013) ﬁnd that ﬁscal consolidation implies an increase in long-term unemployment of about 0.5 percentage points in the medium-term.
Based on a panel of OECD countries.
Other streams of literature study the eﬀects of tax policies and tax shifts on employment.
While a consensus has not been reached, it is unusual for the expenditure side of the budget to be compared to the revenue side of the budget. A detailed review is beyond the scope of my paper, but I would like to note the following: i) Concerning the eﬀects of labor taxation on (cost of) employment, usually based on aggregate data and econometric methodologies, some researchers ﬁnd an eﬀect (Alesina-Perotti, 1997, Blanchard-Wolfers, 2000), while others do not (Bean, 1994);
furthermore, others report mixed results (Daveri et al. (2000) claim that there is a signiﬁcant eﬀect in Continental Europe only). ii) While a signiﬁcant share of public ﬁnance literature highlights the equivalence between consumption and labor income taxes (described e.g. in Auerbach (2006)), others claim the opposite (Blumkin et al., 2012, Sumpson, 1986). iii) Regarding tax shifts, in particular, microsimulations also show an ambiguous picture, Thomas-Picos-Sanchez (2012) claim that a shift from social security contribution to consumption taxes only slightly increases hours, which is not in line with Pestel-Sommer (2013). iv) As regards other eﬀects (growth, eﬃciency, inequality, reform implementation), see among others Auerbach (2006) or Pestel-Sommer (2013).