In this paragraph, we focus on the longitudinal analysis (Cheng et al.2010) of the two cohorts of temporarily laid-off workers, represented by the beneficiaries of Wage compensation fund in derogation (Cassa Integrazione in deroga). The first consists of 135,000 workers who receive the treatment in 2009, while the second is represented by 209,000 workers that receive the treatment in 2010.
The results are shown in Table1. The first panel is relative to the 2009 bene- ficiaries, while the second refers to the 2010 beneficiaries. The share of workers that after the first treatment disappear from the system is reported in the first column (A). The absence of these workers from the system means that they start working.
The second and third column report the share of workers that remain in the same status after 12 months. Hence, they continue to receive the same type of treatment. In this case, we distinguish between Cassa Integrazione in deroga (B), Cassa integrazione straordinaria (Wage compensation special fund) and other forms with unemployment benefits (C) (art.19 of 2/2009 Law).
The other three columns refer to the share of workers that, after 12 months from the first treatment, receive an unemployment benefit. In particular way (D) reports the share of workers in Mobilità in deroga, (E) is the share of workers in Mobilità ordinaria (ordinary Mobility allowance) and finally (F) shows the share of workers that receive an ordinary unemployment benefit (Indennità di disoccupazione). These three categories collect those workers that can be considered formally laid off.
The number of laid off workers is important since it provides a further element to assess the effective rule of theCassa Integrazione in deroga. As a matter of fact, the dismissal represents a failure of this measure in preserving jobs.
Furthermore, the division of the results according with the number of hours of theCassa Integrazione permits to highlight the validity of the STW schemes, in relation to the different type of the entrepreneur’s crisis.
Despite of the progressive worsening of the crisis, the share of workers in CIGD no longer present in the archive SIP during the 12 months after the first treatment raises in the biennium 2009–2010 from 36.4 to 42.6 %.
The share of beneficiaries of a further temporarily laid-off treatment decreases from 49.8 % in 2009 to 42.7 % in 2010 (B?C?D). The share of dismissed workers raises of one percentage point, passing from 13.6 % to 14.6 % (E?F?G).
Table 1 Longitudinal analysis on CIGD workers who complete the first treatment in 2009 and in 2010. Percentage values for status at 12 months
Duration classes of CIGD in hours
CIGD workers who complete the first treatment in 2009. Percentage values for status at 12 months (N = 135.061)
No longer present in the SIP (A)
Re-suspended Unemployment allowance
Total
%
B C D E F G
100 % (zero hours) 25.39 37.81 11.78 1.45 14.06 2.45 7.06 100
From 75 to 99 % 32.77 35.23 9.29 2.71 14.92 2.18 2.9 100
From 50 to 74 % 34.49 40.26 8.29 3.59 10.99 1.19 1.19 100
Form 25 to 49 % 41.2 39.91 7.06 3.96 6.79 0.68 0.41 100
No more than 24 % 49.31 36.59 6.69 3.23 3.53 0.4 0.25 100
Total cohort 36.46 38.22 8.62 3.02 9.92 1.36 2.4 100
Duration classes of CIGD in hours
CIGD workers who complete the first treatment in 2010. Percentage values for status at 12 months (N=209.923)
No longer present in the SIP (A)
Re-suspended Unemployment allowance
Total
%
B C D E F G
100 % (zero hours) 27.14 28.91 13.86 0.73 11.63 2.70 15.03 100 From 75 to 99 % 36.09 28.37 11.94 1.44 14.67 2.79 4.69 100 From 50 to 74 % 40.67 29.55 11.47 3.44 11.27 1.50 2.10 100
Form 25 to 49 % 46.97 31.03 8.84 4.74 6.61 0.85 0.95 100
No more than 24 % 55.95 26.96 6.88 5.50 3.70 0.57 0.43 100
Total cohort 42.67 29.03 10.22 3.47 8.76 1.52 4.33 100
SourceItalia Lavoro and ISFOL on data INPS and Ministero del Lavoro Legend
A Workers no longer in the SIP (reintegrated)
BWorkers present in SIP with a new treatment of Wage compensation fund in derogation (CIGD) CWorkers in SIP with a new treatment of Wage compensation fund not in derogation (CIGS) DWorkers present in SIP with other treatments of suspension
EWorkers present in SIP with an unemployment Mobility in derogation allowance (Mobilità in deroga)
FWorkers present in SIP with unemployment ordinary mobility allowance (Mobilità ordinaria) GWorkers present in SIP with an unemployment allowance (Indennità di disoccupazione)
58 G. De Blasio et al.
Although this percentage difference is slightly low, it is interesting to observe how the recent crisis determines an increase in the share of reinstated workers and a decrease in the share of suspended workers, while the percentage of laid-off workers remains almost the same.
These evidences seem to confirm that investments in active policy measures realized in the context of the 2009 Regional Agreement have encouraged a correct utilization of CIGD, preserving firms from collective dismissal and favouring workers’ reinstatement.
Obviously the duration of the wage compensation fund in derogation (i.e., hours of ‘‘suspension’’ referred to workers involved) affects the results: the lower the number of suspension hours, the higher the likelihood of return to the company.
For workers at ‘‘zero hour’’ (those most at risk of dismissal for companies in crisis) the percentage of those being resettled grows from 25.4 % of the cohort in 2009 to 27.1 % of that of 2010 and the share of reinserted is clearly higher in all the different classes of duration.
These preliminary results do not allow to estimate the causal effect of active policies, and this aspect remains an area of further research. This could be possible through:
• the acquisition of the Regional labour informative systems that allow to access individual data on CIGD workers’ participation to policy measures and analyze the effects of different treatments on the changes of employment status;
• the adoption of counterfactual models. In particular way, a control group could be represented by those CIDG beneficiaries that do not take part to active policy measures.
The second part of this paper regards the analysis of workers who are laid off and receive the Mobility allowance.
In this case, we distinguish between the beneficiaries of the treatment in
‘‘derogation’’ (Mobilità in deroga) from those who receive an ‘‘ordinary’’ treat- ment (Mobilità ordinaria). We analyze the employment outcomes of these two groups in the 24 months after the beginning of the treatment period.
Table2shows a summary of the results. More specifically, 50.4 % of the first group does not receive any job contract, while 49.6 % obtains a job. In particular way, among those who have a job, 17.8 % receives a permanent contract. These results seem to be interesting since they are observed during the peak of the recent economic downturn.
On the other hand, the second group is related to the beneficiaries of the ordinary Mobility allowance (Mobilità ordinaria)—that are not included in the 2009 Regional Agreement. We observe, in this case, that the share of participants who receive at least a job contract is 44.9 %.
Hence, it seems evident that the probability to obtain a job contract is higher for beneficiaries of the treatment in a derogation regime rather than those in an ordinary regime. Analogous considerations are possible when we take into account the type of contract. As a matter of fact, the share of participants who receive a
permanent job contract is higher for beneficiaries of treatment in law derogation.
These results could be due to a different impact of the two policy measures considered or they can be the consequence of systematically different character- istics between firms and beneficiaries in the two groups. In order to assess the roles of the policy measures, in the next section we propose a counterfactual approach.
1.1.1 A Counterfactual Approach
The previous analysis show different results in terms of employment outcomes for beneficiaries of Mobility in derogation with respect to those who receive an ordinary treatment. In order to assess whether these differences are due to a dif- ferent impact of the two policy measures or generated by systematically different characteristics of the subjects involved—firms and beneficiaries, in this section we adopt a counterfactual approach. Our data do not allow to distinguish among firms, while some characteristics for beneficiaries are available. In this regard, this article leaves several discussions as open. A more careful analysis should also investigate firms’ heterogeneity.
The basic idea is to identify for each individual in one group a matching individual from the other that shares similar characteristics. The mean effect of the treatment is then given by the average difference in outcomes between the two Table 2 Longitudinal analysis on workers who start the first treatment in mobility in 2009.
Percentage of workers who sign at least an employment contract in 24 months after treatment Mobilità
in deroga
Mobilità ordinaria
v.a v. % v.a v. %
Workers who start the first treatment in mobility in 2009 18.925 100.0 62.747 100.0 Workers who do not sign any contract of employment within
24 months after treatment
9.534 50.4 34.565 55.1 Workers who sign at least an employment contract within
24 months after treatment
9.391 49.6 28.182 44.9 Workers who sign at least a permanent employment contract
(permanent job) within 24 months after treatment
1.676 8.9 3.250 5.2 Workers who sign at least a temporary employment contract
within 24 months after treatment
7.715 40.8 24.932 39.7 Workers who sign at least a temporary employment contract
transformed in permanent contract within 24 months after treatment
1.196 6.3 5.939 9.5
Workers who sign at least a temporary employment contract not transformed in permanent contract within 24 months after treatment
6.519 34.4 180.993 30.3
Average waiting days for workers who sign at least an employment contract within 24 months after treatment
0.278 0.226
SourceItalia Lavoro and ISFOL on data INPS and Ministero del Lavoro
60 G. De Blasio et al.
groups. In our specific case, we consider beneficiaries of Mobilità in derogation as treated individuals, while the non-treated ones are represented by the ordinary beneficiaries. In order to identify the matching individual, we proceed by esti- mating a probit model as follows:
Pðyẳ1=xị ẳGðb0ỵbxị
where y=1 if the individual is a beneficiary in derogation and 0 otherwise.Gis the standard normal cumulative distribution function, the covariates are represented by the age, region and gender. Our data do not contain other information about indi- viduals, and this represents an element of weakness of this analysis. However, this article would just show how administrative data could be useful for policy evalu- ation purposes. In this sense, a next step should be represented by the integration of different data sources, which could allow to define a more accurate dataset.
Then we estimate the propensity score (Rosenbaum and Rubin, 1983) which represents, in our case, the probability of participation to the Mobilità in dero- gation. This index allows to select for each treated individual the more similar non- treated subject. Table3 reports the mean treatment effect of this counterfactual exercise. The first two columns report the average of the outcome variables for the two groups, while the third column represents the average of the same variables for the counterfactual group, defined on the approach above described. Last column reports the difference in average between the two groups.
As it seems evident, the results confirm that the probability of being employed is higher for treated individuals (Mobilità in deroga). This evidence was confirmed in the descriptive part. On the other hand, the incidence of permanent employees on the whole group of beneficiaries and the incidence of permanent employees on the number of individuals who found a job after the treatment show that the probability to find a more stable job is higher for non-treated individuals. This result is significantly different from that found in the descriptive analysis. All differences result statistically significant.
Table 3 Counterfactual approach
Sample Counterfactual group
Deroga (1) Ordinaria (2) Ordinaria (3) Diff (1–3)
% employed 49.6 44.9 41.9 7.7a
% permanent/total 8.9 5.2 9.6 -0.8b
% permanent/employed 17.8 11.5 22.9 -5.1a
Level of significancea0.01;b0.05
SourceItalia Lavoro on data INPS and Ministero del Lavoro
2 Survival Analysis: An Application of the Kaplan–Meier Model to the Mobility Allowance Beneficiaries
To deepen the descriptive analysis, we propose in this paragraph an application of the Kaplan–Meier survival analysis approach to estimate the probability of exit from the unemployment status for the Mobility allowance beneficiaries.
A previous application of survival analysis models with Kaplan–Meier method was carried out by the Region of Sardinia and Italia Lavoro in the analysis of the customers of the centers for employment services.
The study aims at calculating the probability of the unemployed members to leave the administrative unemployment status. In this paper, we apply the Kaplan–
Meier method to the two cohorts of beneficiaries of the ordinary Mobility allowance and the Mobility allowance in derogation. The analysis is conducted on 2009 beneficiaries and allows to estimate the probability to leave the unemploy- ment status during the 24 months after the beginning of the treatment. The linkage of the two administrative sources permits to estimate this probability, differenti- ating by different social groups. Figure1 shows the probability of survival between the two forms of treatment (ordinary and in derogation). The results confirm hat beneficiaries of Mobility in derogation are characterized by a higher probability to find a (dependent or semi-subordinate) job.
On the other hand, Fig.2shows the different survival curves for beneficiaries of Mobilità, according to their geographical origin. In this case, it is possible to observe an evident difference in the employment probability. As a matter of fact,
Fig. 1 Survival distribution with Kaplan–Meier method for workers who start in 2009 the Mobility allowance treatment (in derogation and ordinary). Probability to sign at least an employment contract in 24 consecutive monthsSourceItalia Lavoro on data INPS and Ministero del Lavoro
62 G. De Blasio et al.
only 30 % of the beneficiaries in the North East area remains in the initial con- dition (unemployed), while the most part could find a job. These probabilities are significantly lower when we consider participants from the South.
3 Conclusion
The results in this paper reveal some interesting aspects that need a further investigation. The use of counterfactual approach (ISTAT,2011) is necessary in order to assess the causal effect of the policy measures adopted. Here, we intro- duced a preliminary counterfactual exercise with the aim to show how adminis- trative data could represent a powerful tool for the analysis of policy evaluation. In this regard, the access to individual data of participants in various policy measures could increase highly the informative capacity of the analysis.
The introduction of Cox regression models (Martin Bland et al.1998), in the survival analysis approach, could be useful in order to establish the factors that may contribute to determine the change of status (age, sex, sector, etc.).
In conclusion, although a number of areas for further research remains and seems necessary, this paper can be considered as one of the first attempts to valorise administrative data sources in the analysis of policy measure impact. We believe that future studies on this topic could not more leave aside from this type of data.
Fig. 2 Survival distribution with Kaplan–Meier method for workers who start in 2009 the mobility allowance treatment (in derogation and ordinary). Probability to sign at least an employment contract in 24 consecutive months for North West, North East, Central and Southern ItalySourceItalia Lavoro on data INPS and Ministero del Lavoro
References
Cheng J, Edwards LJ, Maldonado-Molina M, Komro KA, Muller KE (2010) Real longitudinal data analysis for real people: building a good enough mixed model in statistics in medicine.
Implementation Sci 29(4):504–520, 20 Feb 2010
De Blasio G, Sorcioni M (2012) Employment outcomes of short-time work scheme and unemployment insurance program beneficiaries: a longitudinal approach. Mimeo, Italia Lavoro S.P.A.
ISTAT (2011) INPS, Ministero del LavoroRapporto sulla Coesione sociale. Nota metodologica vol 2. Febbraio
Italia Lavoro (2011)—Regione Sardegna.La transizione verso il mercato del lavoro dipendente dei disoccupati iscritti ai Centri Servizi Lavoro in Sardegna- p 29–31 Ministero del lavoro Luglio
Italia Lavoro (2012) ISFOL, INPS. Rapporto di Monitoraggio: le misure di contrasto alla crisi occupazionale connesse all’Accordo Stato Regioni del Febbraio 2009. Ministero del Lavoro e delle Politiche Sociali—Maggio
Martin Bland J, Altman DG (1998) Statistics notes survival probabilities (the Kaplan-Meier method). Brit Medical J. 317(7172):1572–1580 5 Dec
OECD (2012) The role of policies for labour market resilience. Final report
Rosenbaum PR, Rubin DB (1983) The central role of the propensity score in observational studies for causal effects. Biometrika 70:41–55
64 G. De Blasio et al.
Imbalances, Tensions and Possible Readjustments: Evidence from Intertemporal Accounting and the Financial Accounts
In Need of Sectoral and Regional
Rebalancing in the Euro Area: A Euro Area Sectoral Accounts (Flow-of-Funds) Perspective
Philippe de Rougemont
Abstract The paper looks at the debt crisis through the lenses of the Euro area accounts. It starts from the traditional analysis of the sectoral’s financial balances (savings minus investments), describing their ‘‘rotation’’ during the boom, the crisis and the recovery. It then moves on to emphasise the regional differences in sectoral balances, distinguishing two regional groupings: external surplus and deficit countries. The boom period is mostly marked by a pronounced swing into deficit of the private sector in deficit countries, which contrasts with the fairly stable private sector surpluses in surplus countries, resulting in a widening gap in external balances between the two groups. However, the abrupt reversal in private sector financial balances in the deficit group after Lehman hardly changed the current account surplus/deficit configuration of these two groupings over 2008–2011, but instead was largely compensated by a larger gap in government deficits. Taking another perspective, the pre-crisis gradual widening gap in external balances between these two groupings can be seen as originating largely from increasingly large differentials in national saving, rather than in national investment. This increasing national saving differential mostly reflected increased saving differentials of the corporate sector up to 2007 (rather than households’ or governments’) that have reversed only to a limited extent since then. This in turn resulted from the emergence of a considerable gap in gross operating margin between the two regional groupings, with much higher margins in surplus coun- tries. This reflected the faster increase in wages in the deficit countries compared to surplus countries, in excess of what would be justified by productivity and/or
The views expressed are those of the presenter and should therefore not be reported as representing the views of the European Central Bank.
P. de Rougemont (&)
European Central Bank, Frankfurt am Main, Germany e-mail: Philippe.de_Rougemont@ecb.int
L. Paganetto (ed.),Public Debt, Global Governance and Economic Dynamism, DOI: 10.1007/978-88-470-5331-1_6,Springer-Verlag Italia 2013
67
growth differentials. Our analysis suggests that the very large wage gap (in the order of 15–17 %) that had emerged would need to be reduced as a precondition to macro-rebalancing. The paper highlights the mechanism by which the free cir- culation of savings in a (financially integrated) monetary union distorts price/wage structures, in the absence of fully integrated goods and labour markets. This contrasts with a more optimistic interpretation that suggested that greater circu- lation of savings within the euro area was a welcome consequence of increased financial integration, celebrating the end of the Feldstein-Horioka puzzle. The distortion of the relative price/wage structure, away from initial equilibrium, is in itself not a difficulty in a common currency area, unless the lack of price/wage flexibility prevents, once capital inflows stop or even starts reversing, a rapid return of prices and wages to their original equilibrium level.
1 Introduction
Persistent intra-euro area current account as well as sectoral imbalances had been building up in the years prior to the financial turmoil. This paper aims at providing a closer look at regional imbalances and heterogeneities in the run-up to the financial crisis and in more recent quarters, drawing on the country information underlying the sectoral euro area accounts (flow of funds). This paper builds on data and analyses already presented in Box 3 of the February 2012 Monthly Bulletin of the ECB as well as in the ECB Financial Integration Report dated April 2012 (Feature E).1
The paper looks at the evolution of the crisis through the integrated and con- sistent lens of the Euro Area Accounts (EAA, see Box), which brings together the financial and non-financial accounts of the different institutional sectors (i.e.
households, non-financial corporations, financial corporations and general gov- ernment) and the rest of the world, presenting data in nominal rather than real terms. Having consistent flows and balance sheets makes it easier to analyse the accumulation of imbalances and associated balance sheet vulnerabilities.
The paper continues with discussing the evolution of the net lending/net bor- rowing by sector, the ‘bread and butter’ of the flow of funds analysis, in the run-up to the boom, during the 2008–2009 recession and then the recovery. As a com- plement to the euro area perspective, we add a regional perspective of the sectoral net lending/net borrowing—by grouping countries according to whether they had a current account deficit or surplus in the pre-crisis period (5 years)–, to better understand the sectoral pattern behind the widening gap in current accounts between country groups.
1 See also the article ‘‘The financial crisis in the light of the euro area accounts’’ (ECB2011a).