4.1. Descriptives
Firstly, we can look at the size of the companies analysed. The criteria established by the European Union to classify companies based on size were used (Commission Recommendation of 6 May 2003. Available online:
https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32003H0361, accessed on 16 April 2021). Companies with fewer than ten employees and an annual turnover or annual balance sheet total not exceeding EUR 2 million are considered micro-enterprises. A small company is one with 10–49 employees and an annual turnover or annual balance sheet total not exceeding EUR 10 million. Medium-sized companies have 50–249 employees and an annual turnover not exceeding EUR 50 million or an annual balance sheet total not exceeding EUR 43 million. Finally, large companies have more than 250 employees and an annual turnover exceeding EUR 50 million or an annual balance sheet total exceeding EUR 43 million.
As shown in
Figure 1, most of the sheltered employment centres operating in Spain are micro-companies. Small enterprises represented 30% of the total sample, accounting for 300 sheltered employment centres. In total, 217 companies analysed were medium-sized companies, and lastly, there were 59 large companies, around 6% of our sample.
The number of employees ranges from micro-companies with only one worker up to one company with more than 7000 employees. On average, they have 74 employees, although we must remember that the standard deviation for this variable is relatively high. Another vital piece of data when analysing sheltered employment centres is their age. On average, these companies have been operating for almost 19 years, which means that they have experience in the market. The descriptive results for the age and number of employees variables are shown in
Table 2.
Secondly, looking at the legal structure of sheltered employment centres (see
Figure 2), the results show that 88% of the sample are limited partnerships. The main difference between a limited company and other types of companies is the fact that the liability of its members is limited to the capital contributed, i.e., the owners are not obliged to pay the company’s debts out of their own capital.
Next, the location and sector were analysed. We use the sector categories from the NACE (Statistical Classification of Economic Activities in the European Community) (see
Table 3). In Spain, there are 17 autonomous regions, and there is a different regulation about grants and subsidies for the sheltered employment centres for each autonomous region. Therefore, the variable location is important because it can condition the subsidies received by the sheltered employment centres.
Most of the sheltered employment centres are in Andalusia, Madrid, Valencia and Catalonia. However, these do not account for all sheltered employment centres in Spain. As mentioned previously, there is no single public list of all the centres. In terms of the sector of activity of these companies, most of them operate in the service sector (306 companies provide administrative activities and auxiliary services), around 30% of the total sample. This could be due to many medium and large companies outsourcing this kind of service.
Additionally, these businesses are involved in accessible and very standardised activities that can be performed with less skill. The wholesale and retail trade, vehicle repair, transport and storage (170), together with different industry activities (food, textile and manufacturing) (176) are the activities most performed by these companies. Lastly, 103 centres are involved in health and social work activities.
The descriptive analysis of the ratios is based on the last five years so that we can see the evolution during this period. As shown in
Table 4, the trend for the ratios significantly improves during the period analysed, by 50% comparing the positions in 2015 and 2019. The indebtedness and solvency ratios are relatively stable, particularly the second of these. The profitability analysis, completed using ROA and return on equity (ROE) ratios, shows positive values over the period. ROA in 2019 is almost double the figure in 2015. Although sales growth is significantly lower in 2019 than in 2018, it is crucial to highlight that sales per employee did not fall by the same proportion, which shows an improvement in the efficiency of human resources. This is in line with previous studies [
64]. The financial risk is relatively low in the period, mainly in the later years. The cash flow resulting from the relationship with the government (this being the difference between corporation tax paid and capital subsidies received) shows that the change from a negative relationship in 2015 to a very positive one in 2019 is favourable for the government. This corroborates the idea that these businesses are sustainable from a public economic policy perspective and of benefit to society.
Table 5 shows the evolution of the source of income for the sheltered employment centres in our sample. The trend for private funding increased from 2015 to 2019. At the same time, public funding fell significantly, by over 60%. In addition, other operating income increased during the period analysed, with the same trend in the sales of products and services. Of the EUR 8446.99 thousand of total annual incomes, 92.9% comes from sales of products and services; 0.39% from public administration subsidies; and only 1.72% comes from private subsidies; and 17.19% from other operating income in 2019. Results in previous years are in line with 2019 results.
As shown in
Table 5, sheltered employment centres contributed more to the public finances than they received. Therefore, the balance is positive for the government. Consequently, encouraging this type of social entrepreneurship can contribute to the welfare of disabled people but can also contribute to society’s welfare in general because it can generate net income for the State.
Analysing the ratios and variables such as age and number of employees, we see that these companies are socially sustainable. For the number of employees, it is evident that these companies contribute to social sustainability as they generate employment for people with disabilities. It should be recalled that at least 70% of staff in these companies must have a degree of disability equal to or greater than 33%. Thus, sheltered employment centres directly promote and generate social sustainability. Moreover, if we analyse the age of these companies, we can see that they have experience in the market. They have been operating for many years, and this creates more of a guarantee of employment for people with disabilities. At the same time, they can generate new jobs.
If we analyse the economic sustainability achieved by sheltered employment centres based on the descriptive analysis, we see that they also fulfil this function. The profitability ratios show that they do not earn high profitability but achieve the minimum based on their activities. However, the ratio that measures the return to shareholders is quite high. In addition, the sales ratios show that these companies are economically sustainable in their core activities. Therefore, we can say that these companies actively contribute to social and economic sustainability [
49,
51].
4.2. PART Algorithm
To analyse the relationship between the selected independent variables and financial profitability (ROE), we developed a rule induction algorithm, specifically a PART algorithm. A PART algorithm [
65] is a classifier that generates production rules by incorporating a modified form of the C4.5 decision tree and eliminating some of the paths found in an initial decision tree structure. Therefore, PART is a rule induction algorithm based on partial decision trees [
66]. The main advantage of this classifier is its simplicity, as it develops the most general rule by choosing the leaf that represents the most significant number of situations. The rules obtained are logical sentences of the form ‘if conditions then decisions’, that is, the rules specify what decisions (actions) should be undertaken when some conditions are satisfied.
We have chosen ROE as the dependent variable because it measures the return on the capital provided by investors in terms of the net profit obtained in the year. It is an indicator of both economic and social return. The ROE variable has been codified as 1 for a positive return and 0 otherwise, to develop our model.
ROE is an important ratio for investors, allowing them to understand the creation of additional value [
67], but also for entrepreneurs and managers because it helps them in the decision-making process [
68,
69,
70,
71]. With ROE, investors can assess if their investment is profitable or not, analysing the global efficiency rate [
72]. We use ROE in this study because the government could be considered another “shareholder” when subsidies are paid in the same way as shareholders deposit their contributions.
To validate the rules, we have used a cross-validation procedure obtaining an 85.10% share of correctly classified sheltered employment centres (that is, accuracy). The satisfactory results in terms of accuracy allow us to interpret the strongest rules (see
Table 6) to draw the following conclusions. First, there are more rules for class 1 than for class 0. This result indicates that it is easier to classify profitable firms (class 1) than unprofitable ones (class 0). There could be many different causes of unprofitability. Second, the key ratios to classify unprofitable companies would be liquidity and solvency (which would take negative or very low values) together with sales growth in this type of company. This means that this kind of firm is very conditioned by short-term ratios. Since they are mainly operating in the sector providing administrative activities and auxiliary services, they are dependent on cash. Third, the common ratios associated with solvent companies are the financial risk and indebtedness ratios. These ratios are related to the method of financing. They are more solvent if they are externally financed via banks. The reason is that banks only finance those businesses which are solid and have a market strategy and good leverage. A non-viable business does not receive funding from banks because financial institutions need guarantees about solvency and the repayment of the loans granted. Logically, if the companies have more debt, they have higher financial expenses. Thus, the two ratios (F-risk and Indb) are related. Sales per employee is also an important ratio for viable sheltered employment centres. As they are very labour intensive and have a social function to hire as many disabled workers as possible, the sales per employee ratio should be as high as possible (more than 11.5 in the case of the third rule, with 184 positive cases but 14 mistakes). Finally, according to the fourth rule, large service companies with less external financing are also profitable (178 positive cases against 41). These results are in line with the results of the descriptive analysis, which reinforces the argument about the economic sustainability of sheltered employment centres.