• employer
• job quality
• knowledge
• occupational groups
• power
• safety risk
• virtuous dynamics
• workplace innovation

Technological progress brings about substantial changes in the structure of employment. For a long time, economic research has focused on analyzing the relationship between innovation and job creation and destruction, completely neglecting the qualitative dimension of work. However, the quality of work positions is crucial for increasing labor force participation, well-being and economic performance.
Which occupations are more exposed to workplace innovation? High-quality or low-quality ones? This thesis aims to address these latter questions.
The empirical analysis is developed by merging two datasets at the level of occupational categories (ISCO 4-digit classification), namely the 2013 ICP (Indagine Campionaria sulle Professioni) and 2018 PLUS (Participation, Labour, Unemployment, Survey) sample surveys produced by the National Institute for Public Policy Analysis (Inapp).
The first dataset was used to construct an index of job quality (JQI), necessary since, so far, there is no single methodology or comparable metric established for its assessment. Therefore, starting from the OECD Job Quality Framework, a framework was defined to comprehensively account for i) the working environment, ii) employment conditions and iii) employee empowerment. By means of a factor analysis, five main drivers were identified for conceptualizing and operationalizing occupational well-being by occupational classification, namely i) the degree of knowledge required and the learning regimes to which workers are called upon to perform work, ii) their right to participate in the work process, thus their authority to intervene, iii) the characteristics of the work environment concerning relations with third parties, iv) the degree of safety risk and v) the level of psychological and/or material benefits related to the profession, such as salary satisfaction, recognition of colleagues or possibilities for career advancement. Hence, the JQI was calculated for every i occupational class as the summation of the score of each of the five factors associated with the i-th ISCO 4-digit occupational code; then, the index was normalized within a range between 0 and 1.
Once this metric was defined, a probit regression model was performed to define how much an individual’s probability of receiving an innovation depends on the level of quality of the work he/she performs in his/her occupational category. In other words, whether the quality of work preceding innovation is a good predictor of the occurrence of a technological change in the workplace. Note that the ‘Innovation’ variable, obtained from the 2018 PLUS survey, was derived from the response of the interviewed units to the question "Have any relevant technological innovations been introduced in your workplace in the last two years (2016-2017)?", where technological innovation is defined as the introduction of new products and services, as well as new ways of producing, distributing and using them, such as, for example, the introduction of electronic/digital devices like tablets or smartphones, which have replaced analogue and/or physical devices previously used to perform tasks in the organization where one works. It should also be stressed that the probit performed considers several control variables in order to take into account the specific characteristics of both the individual interviewed and the enterprise to which the latter belongs, as they may influence the likelihood of experiencing an innovation.
The empirical analysis reports that professional groups with higher levels of occupational well-being are more likely to experience innovation in the workplace. Furthermore, to limit, as far as possible, the problems of selection bias that might arise insofar as other confounding factors might explain the sorting into high- or low-quality occupations, in turn influencing the probability of innovation, a propensity score matching was carried out. The approach used for the implementation of this statistical technique was the 1-to-1 nearest neighbour matching method. Treated and untreated individuals were distinguished based on the level of their JQI. Setting a threshold value of 0.54, all workers with a JQI below this value were considered untreated, while those with an index above the threshold as treated.
The average treatment effect (ATE), i.e. the estimate of the expected effect on the outcome if the individuals in the population were randomly assigned to the treatment, confirmed a positive and statistically significant relationship between the variables JQI and Innovation, albeit less strong.
These results have led to some considerations. It was pointed out that a good quality of work can itself be an engine of innovation; indeed, the greatest number of technological changes between 2016 and 2017 were recorded among the occupational groups with the highest job quality indices in 2013. Furthermore, according to several scholars, the adoption of more technically advanced production processes and/or equipment leads itself to an increase in the quality of work, as it often requires an enhancement of learning practices, coordination, teamwork and discretionary power, all key factors of the job quality index developed in this thesis. Nevertheless, it was not possible to probe this consideration in this analysis, as the data used to calculate the JQI precede the innovation data.
It is also necessary to look at the other side of the coin, at the negative reflex to which these results lead, namely that low quality jobs remain as such without receiving innovation. Technological changes occurred among the occupational groups that needed it the least, namely those who already had ex ante a higher level of well-being at work, not affecting employees worse off, especially in terms of environmental safety and occupational health.
To summarize, the total result is on the one hand positive in that it can positively enhance virtuous dynamics between innovation and quality of work, and on the other hand negative in that innovation does not seem to be triggered in cases where the quality of work is lower, amplifying differences between professional groups.
A further reflection arising from the results is that the employer plays a key role in shaping the JQI. In fact, it is the employer who chooses the level of investment in R&D, training and implementation of occupational health and safety practices and defines the organizational design from which the level of knowledge required and accumulated, autonomy and discretion exercised, and co-operation are derived. Unfortunately, this aspect could not be investigated in this study due to lack of data, but it is suggested for future research to link the JQI variable to the company, including variables such as technology, dynamic innovation at the product and process level, and the level of investment as proxies for the role of the employer.