Regional Performance Trends in Providing Employment for Persons with Disabilities: Evidence from Italy

Abstract

The aim of this work is to analyse the situation of persons with disabilities in the Italian labour market, with a view to providing guidelines to promote their inclusion both in the labour market and in society. For this purpose, we propose a two-step analysis focusing on Italian regions for the period 2006–2009. In the first phase, we use the Data Envelopment Analysis method to evaluate regional efficiency in providing employment for persons with disabilities. Cluster analysis is then applied to regional efficiency scores and economic policy variables in order to identify “policy clusters of regions”. Our results show that it is necessary both to focus on the residual work ability of persons with disabilities and to develop a social integration culture on the demand side of the labour market. Moreover, a structural reform of disability benefit systems is required in order to promote a culture of social inclusion.

Appendices

Appendix: The Method of Penalty Coefficient of Variation: The Mazziotta-Pareto Index (MPI)

The method of the penalty coefficient of variation allows us to construct a composite measure of the infrastructural equipment of a set of territorial units, assuming that each component is not interchangeable, or only partially, with the other.¹² In this context, the aggregate function (arithmetic mean of the standardised values) is corrected by a penalty coefficient that depends, for each territorial unit, on the indicator variability with respect to the average value (“horizontal variability”). This variability, measured by the coefficient of variation, penalises the score of each unit which, showing the same mean value, displays a greater disequilibrium among the indicator values.

Finally, the use of standardised deviations allows us to obtain a “robust” measure, less influenced by outliers (Mazziotta et al. 2010). This approach needs a balanced endowment of elementary components (Mazziotta 2005; Mazziotta and Pareto 2007; Mazziotta et al. 2010).

Construction of the indicator proceeds in the following stages.

Normalisation of Indicators

Let X = {x _ij } be the matrix with n rows (geographical units) and m columns (indicators) and let M _xj and S _xj denote the mean and the standard deviation of the j-th indicator:

$$ M_{{x_{j} }} = \frac{{\sum\nolimits_{i = 1}^{n} {x_{ij} } }}{n},S_{{x_{j} }} = \sqrt {\frac{{\left( {\sum\nolimits_{i = 1}^{n} {x_{ij} - M_{{x_{j} }} } } \right)^{2} }}{n}} $$

The standardised matrix Z = {zij} is defined as follows:

$$ z_{ij} = 100 \pm \frac{{\left( {x_{ij} - M_{{x_{j} }} } \right)}}{{S_{{x_{j} }} }}10 $$

In this type of normalisation the “ideal vector” is the set of mean values and it is easy to identify both the units that are above the mean (value greater than 100) and the units that are below the mean (value less than 100) (De Muro et al. 2011).

The sign ± depends on the relation of the j-th indicator with the phenomenon to be measured, such as the quality of work (+ if the individual indicator represents a dimension considered positive and—if it represents a dimension considered negative). In our case, since we want to calculate a quality of work indicator¹³, the j-th indicator of the unemployment rate will have a negative sign, while the j-th indicator of the youth employment rate will have a positive sign. We will apply the same logic to social capital indicators (bonding, bridging and linking social capital).

Aggregation

Let CV = {cv _i } be the coefficient of variation for the i-th units:

$$ cv_{i} = \frac{{S_{{z_{i} }} }}{{M_{{z_{i} }} }},\;\;{\text{where}} $$

$$ M_{{z_{i} }} = \frac{{\sum\nolimits_{j = 1}^{m} {z_{ij} } }}{m}\,\,{\text{and}}\,\,S_{{z_{i} }} = \sqrt {\frac{{\left( {\sum\nolimits_{j = 1}^{m} {z_{ij} - M_{{z_{i} }} } } \right)^{2} }}{m}} $$

Construction of the Composite Index

The composite index based on the penalty coefficient of variation method can be written in the following generalised form:

$$ MPI_{i}^{ \pm \, } = M_{{z_{i} }} \pm \, S_{{z_{i} }} cv_{i} $$

where the sign ± of the penalty depends on the kind of phenomenon to be measured and hence on the direction of the individual indicators (De Muro et al. 2011).

If the indicator is “increasing” or “positive”, i.e. increasing values of the indicator correspond to positive variations of the phenomenon (e.g. the quality of a work of a region), then we use the MPI with a negative penalty:

$$ MPI_{i}^{ - \, } = M_{{z_{i} }} - \, S_{{z_{i} }} cv_{i} $$

Vice versa, if the indicator is “decreasing” or “negative”, i.e. increasing values of the indicator correspond to negative variations of the phenomenon (e.g. the poverty of a region), then we use the MPI with a positive penalty:

$$ MPI_{i}^{ + \, } = M_{{z_{i} }} + \, S_{{z_{i} }} cv_{i} $$

In the former example, the penalty coefficient corrects the mean of the standardised indicators by pushing it down, whilst in the latter case it pushes it upwards.

For the quality of work and bridging and linking social capital, the indicator has a negative sign: this means that increasing values of the indicator correspond to positive variations of the quality of work of a region and of these two types of social capital. On the other hand, in the case of bonding social capital, the indicator has a positive sign, as this type of social capital is seen as negative in the literature on social capital (see Sabatini 2005, 2007, 2008, 2009, and Sect. 3 of this work).

In the following tables we list the variables used in the construction of our indicators. We also indicate the sign with which each variable contributes to the building of indicators (see Tables 5, 6, 7, 8).

Table 5 Bonding social capital

Table 6 Linking social capital

Table 7 Bridging social capital

Table 8 Quality of work