DISCUSSION Employment Protection, Workforce Mix and Firm Performance

Any opinions expressed in this paper are those of the author(s) and not those of IZA. Research published in this series may include views on policy, but IZA takes no institutional policy positions. The IZA research network is committed to the IZA Guiding Principles of Research Integrity. in labor and offers evidence-based policy on labor market issues. IZA runs the world’s network of economists, whose research aims to provide answers to the global labor market challenges of our time. Our key objective is to build bridges between academic research, policymakers and society. ABSTRACT We measure the impact of employment protection reduction in an uncertain framework on firms’ hires and performance, exploiting the Italian 2015 Jobs Act. Results indicate that firms (1) stabilize workforce mainly through contract transformations of low-tenure and low-human-capital incumbent workers performing high-physical and low-intellectual tasks; (2) apply a cost-saving strategy that increases profits and decreases value added per-head. Effects are stronger among non-exporting and non-innovative firms. Our evidence casts doubts on the effectiveness of employment protection reductions in enhancing productivity in the long run.


Introduction
Since it became a widely applied policy in the early nineties, labour market deregulation was implemented by relaxing the conditions to hire under temporary contracts (Berton et al., 2012), thus exacerbating the issue of dual labour markets (Hijzen et al., 2017). To reequilibrate the gap and foster permanent employment, several reforms were approved after the 2009 crisis (Eichhorst et al., 2017). Italy epitomizes this narrative. In the present study we perform a mid-term diff-in-diffs evaluation of its Jobs Act, which reduced employment protection (EPL) for open-ended contracts signed in firms with more than 15 employees since March 2015.
Our work adds to the long-lasting debate on the impact of EPL changes in a context of economic uncertainty (Bentolila and Bertola, 1990). We question whether decreasing EPL helps firm performance and through which channel, i.e. enhanced productivity versus cost competition. In particular, we focus on small/medium firms (employing about 20% of workers) and evaluate the impact of an EPL reduction on their gross and net worker flows (Sestito and Viviano, 2018), investigating changes in their workforce human capital (HC) mix as measured by education (Charlot and Malherbet, 2013), skill/task (Kahn, 2018) and previous tenure (an original addition to the existing literature); we then evaluate the impact on firm performance as measured by productivity (Autor et al., 2007) and profits (Bjuggren, 2018) and single out innovative (Griffith and Macartney, 2014) and exporting firms (Selwaness and Zaki, 2019).
We exploit a linked employer-employee dataset of labour market histories merged with firm-level balance-sheets to study the above-mentioned outcomes at the plant/firm level. We study those outcomes jointly as suggested in Boeri (2011)  Our analysis suggests that workforce HC mix downgraded and firms improved their resilience through a cost-saving rather than a productivity-enhancing strategy. This is consistent with surviving in the aftermath of an economic crisis. However, it casts serious doubts upon sustainability in the longer run, in a stagnant-productivity country.

Data
We rely upon the administrative dataset called Comunicazioni Obbligatorie (COB), transformations) at the plant level in 2013-2017. Plants can then be linked to their parent firm. We focus on medium-size firms around the reform threshold (9-30 employees), fully located in Piedmont. 1 The final dataset includes about 16,500 firms (17,000 plants) and it is a de-facto balanced panel, reflecting the low probability of shutting-down for plants/firms employing at least 9 workers.
We complement COB with ASIA, an administrative archive of all active enterprises, to recover firm-size stock measurement, and with AIDA, an archive of -sheets to retrieve information on firm performance.

Models and identification strategy
Our identification relies on a difference-in-differences strategy which exploits the setting of the Jobs Act, that reduced EPL only for firms with more than 15 employees. Following Bjuggren (2018) Euros, respectively).
We assess the effects of the reform by estimating the following equation   (Tables C1-C2) and stronger among non-innovative firms (Tables C3-C4).      (Table C11).

4.1.
Concluding, in 2015 a strengthening strategy was in place, where workers shifted from temporary to open-HC adopting a surviving rather than a value-added enhancement strategy. We now move

Fig. 4: New open-ended contracts by tenure and tasks
Source and Note: see Figure 1

Model 2. Firm performance
Productivity, as measured by value added per worker, shows a decreasing trend in treated ( Figure 5). As employment was increasing (Figure 1), we deduce that value added did not catch up with employment.

Fig. 5: Firm-level productivity
Source and Note: see Figure 1. No s.e. clustering for innovative firms due to small sample size.

Fig. 6: Effect on firm-level performance
Source and Note: see Figure 5 Despite a stagnant/decreasing productivity trend ( Figure 5), Figure 6 indicates that profits did not suffer, particularly in non-innovative and non-exporting firms. We deem this coherent with results on HC: following a reduction in EPL, firms decreased HC penalizing productivity and following a cost-saving strategy. The effect was stronger among non-exporting/non-innovative firms.

Conclusions
has been said regarding their joint relationship with EPL. We assess that an EPL reduction may be unsuccessful in enhancing HC, as small/medium-size firms take advantage of the reform to become more resilient through a defensive strategy rather than by investing. This is mostly true for those not exposed to international competition or not innovative.
External validity is limited, as no evaluation design is available for larger firms; however, 9-30 employee firms cover a significant share of employment in Italy. Such policies should therefore be more carefully designed to trigger the desired growth goals.  derived from external sources to describe more precisely job contents and tasks along the physical and intellectual dimensions and to identify which are the exporting/nonexporting and innovative/non-innovative firms in the sample. What follows briefly describes how these variables have been created and the source of information used.

Physical and Intellectual Tasks
The variables describing physical and intellectual tasks are derived from a set of indicators of task content developed by Eurofound (2016). These indicators are constructed at job cell level, defined by the combination of two-digit occupations (ISCO 08) and sectors (NACE Rev 2.0), with values ranging from 0 to 1. In our analysis, we define a job as characterized by high or low physical/intellectual task if its relevant index is above or below the median computed in the sample.

Exporting / Non-exporting Firms
The indicator variable identifying the Exporting firms is derived by our balance-sheet information. Although it does not include a flag explicitly accounting for firms active on international markets, we recover such piece of information by a word search strategy in firms with plant located outside Italy.