VoxEU Column Labour Markets Welfare state and social Europe

Local interactions between public and private employment

The nature of the relationship between public and private employment is ambiguous, with studies showing that increased public employment can have both crowding-in and crowding-out effects on private employment. This column explores this relationship across Italian municipalities. It finds evidence of strong crowding-out effects across municipalities, which is partially explained by increased competition in the housing market.

The interaction between private and public jobs has always been the subject of hot policy debate. Several countries have used public employment as a way to contain unemployment in backward areas, but these interventions have been harshly criticised because they may crowd out private employment.

Consequences for the local economy

An increase in public employment in one area boosts the demand for local products and, by providing additional public services, it may favour a business-friendly environment or better amenities. A strong local multiplier might therefore imply positive (crowding-in) effects on private employment. On the other hand, more public employment can raise the cost of local production by increasing the price of scarce resources, so as to crowd out the private sector (Farmer and Plotnikov 2011). In a recent paper, we use a model of spatial equilibrium to show that a rise in public employment, by boosting demand and housing prices, may reduce local real wages and induce private-sector workers to move elsewhere (Auricchio et al. 2017).

The evidence on the net impact of public jobs on total employment is mixed. Early papers strongly supported crowding out, but focused only on international comparisons (Algan et al. 2002, Behar and Mok, 2013). More recent research exploiting within-country variation comes to different conclusions. Faggio and Overman (2014), who first looked at local labour markets in the UK, found that the negative impact on the tradable sector is offset by a positive impact on non-tradables. Jofre-Monseny et al. (2016) find an overall positive effect (crowding-in) driven by the non-tradable sector in Spanish cities, while Becker et al. (2015), who analyse the move of the capital from Berlin to Bonn in the wake of the WWII, document basically no impact on private employment.

The Italian case

Italy is an interesting case study. Employment in the public sector has been exploited to redistribute resources from the richer North to the poorer South (Alesina et al. 2001). In Italy, public sector wages tend to be higher than in the private sector (Depalo et al. 2015) and job security is full, since layoffs are limited only to disciplinary matters. However, the sustainability of the cost of the public sector has become a primary concern and, over the last two decades, several laws have imposed a significant contraction on the public workforce. The reduction was mostly implemented through a halt in turnover, by imposing strong limitations to new hiring and relating the changes in employment to the pre-existing level of public employment, with not much relationship to local labour conditions.

Identification strategy

We exploit these strong limitations to turnover in the public sector to study the impact of changes in public employment at the local level on private employment and house prices. We use data from Italian municipalities, looking at the changes between the last two census waves (2001, 2011). Changes in local public employment were strongly influenced by central government decisions, with little attention paid to the economic conditions of the municipalities. This motivates the use of an instrumental variable identification strategy (as in Faggio and Overman 2014). Essentially, we instrument actual changes in public employment by exploiting its incidence and sectoral mix at the beginning of the decade interacted with the changes that occurred at the national level.

We also exploit the variation in the coefficient of interest induced by the introduction of relevant control variables (following Oster 2016), to argue that a reasonable degree of failure of the exclusion restriction does not invalidate our main conclusion. The findings also survive the inclusion of a large set of observable municipality characteristics, selected by machine learning methods from a rich database (Belloni et al. 2014).


Our results strongly support the conclusion that crowding-out occurs. Each additional public worker reduces private employment by nearly 0.7 units, which implies that the net impact on local employment is only around 0.3. The negative effect on the private sector mostly affects the tradable sector (manufacturing), with no impact detected for other sectors. Further estimates suggest that house prices grow in those areas that have been hit by a positive shock of public employment, consistent with the implications of our theoretical framework.

Our main result echoes the evidence obtained from international comparisons, but it contrasts with findings of studies that focus on local labour markets, which show null or even positive overall effects on private employment. One reason for this difference could be the rigidity of local housing markets in Italy, making the displacement which occurs through the housing market channel more relevant. Another possible reason is that, as argued by de Blasio and Menon (2011), local multipliers are quite small in Italy and, therefore, the impact of increased demand on the non-tradable sector is quite limited. We also estimate that public employment growth leads to an increase in the working age population by more than the growth in total employment, suggesting that individuals’ residential choices also depend on perspectives about the local availability of public jobs.


Although public employment might be used as a subsidy for areas lagging behind, so as to limit the population outflow, policymakers should be aware of its potential downsides. In the case of Italy, we highlight that when public employment increases, private employment will decrease, a finding that can be – at least partially – explained by a rise in local house prices. These consequences limit the effectiveness of public employment as a redistributive policy tool. At the same time, our evidence suggests that areas hit by stronger contractions in public employment – such as those caused by recent policies targeting that decreased public sector turnover – seem to have experienced a relative increase in private sector jobs. This finding reduces the strength of one of the main arguments commonly put forward against reforms that aim at contracting the size of public employment.

Authors’ note: The views expressed here are those of the authors and do not necessarily reflect those of the Bank of Italy.


Alesina, A, S Danninger and R Massimo (2001), “Redistribution through public employment: The case of Italy”, IMF Staff Papers 48(3): 447-473.

Algan, Y, P Cahuc and A Zylberberg (2002), “Public employment: Does it increase unemployment?”, Economic Policy 34: 7–66.

Auricchio, M, E Ciani, A Dalmazzo and G de Blasio (2017), “The consequences of public employment: Evidence from Italian municipalities”, Bank of Italy, Working paper no 1125.

Becker, S O, S Heblich and D M Sturm (2015), “The impact of public employment: Evidence from Bonn”.

Behar, A and J Mok (2013), “Does public-sector employment fully crowd out private-sector employment?”, IMF Working paper 13/146.

Belloni, A, V Chernozhukov and C Hansen (2014), “High-dimensional methods and inference on structural and treatment effects”, Journal of Economic Perspectives 28(2): 29-50.

de Blasio, G and C Menon (2011), “Local effects of manufacturing employment growth in Italy”, Giornale degli Economisti 70(3): 101–112.

Depalo, D, G Raffaela and E Papapetrou (2015), “Public–private wage differentials in euro-area countries: Evidence from quantile decomposition analysis”, Empirical Economics 49(3): 985–1015.

Faggio, G and H Overman (2014), “The effect of public sector employment on local labour markets”, Journal of Urban Economics 79(C): 91–107.

Farmer, R and D Plotnikov (2011), “Does fiscal policy matter? Is there a better way to reduce unemployment?”,, 5 September.

Jofre-Monseny, J, J I Silva and J Vázquez-Grenno (2016), “Local labor market effects of public employment”, Document de treball de l’IEB 2016/11, Institut d’Economia de Barcelona.

Oster, E (2016), “Unobservable selection and coefficient stability: Theory and validation”, Journal of Business Economics and Statistics.

1,050 Reads