Precarious work is a widespread phenomenon in both advanced economies and in emerging and developing countries and remains a major policy concern in many contexts. According to a survey conducted by the Ipsos institute on 12,000 adult workers residing in 27 countries (Ipsos 2020), 54% of the sample were concerned about losing their job in the 12 months following the interview (17% “very concerned”, 37% “fairly concerned”). The highest values were recorded in Russia (75%), Spain (73%), and Malaysia (71%), while the lowest ones were reported in Germany (26%), Sweden (30%), the Netherlands, and the United States (36%). Although they indicate strong variability between countries, data demonstrate that job insecurity, defined as the subjectively perceived probability of experiencing an interruption in one's career (Shoss, 2017), is a pervasive issue even in nations with a strong industrial vocation and generous welfare measures (as in the case of Germany and Swede Since the 1990s, global structural shifts loosened traditional worker–firm ties (Davis, 2013).This sparked sustained scholarly interest in job insecurity. Globalization and automation cut costs and cycle times, yet heightened perceived insecurity (Scheve & Slaughter, 2006; Couple, 2019; Nam, 2019; Cao & Song, 2025). Technological progress threatens not only low-skilled workers, but also highly qualified professionals (Colvin, 2015). The diffusion of Artificial Intelligence has amplified these concerns, with similar effects for managerial and non-managerial staff (Koo et al., 2021). AI’s negative impact is partly mediated by vocational learning ability - the capacity to autonomously acquire and apply new skills (Liu & Zhan, 2020). Constant AI interaction pushes continuous upskilling, shaping insecurity, creativity, and well-being (Wu et al., 2024). Fear of rapid skill obsolescence - and of human roles being replaced by virtual ones - adds further stress (Sharif et al., 2025). The approach adopted by researchers to date has favored the analysis of individual and organizational drivers of job insecurity, paying little attention to systemic ones. In particular, they have used survey data relating to variables such as contract type, responsibilities associated with one's position, hourly wage, health conditions, number of children, age, years of experience, HR policies and so on. Such analyses have often focused on specific sectors and regions to account for the heterogeneity present in different economic sectors and in the economic conditions of various countries, proving useful in evaluating the predictors of job insecurity in different contexts (Lee et al., 2018; Chirumbolo et al., 2020; Martínez et al., 2020; De Cuyper et al., 2021; Muñoz Medina et al., 2023; Darvishmotevali, 2025). However, a fundamental limitation of the individual-based approach is its inability to account for macroeconomic variables and (consequently) the influence that the alternation of the various phases of the business cycle exerts on job insecurity. Indeed, while purely individual variables undoubtedly have a significant impact on job insecurity, it is equally plausible that a non-negligible component of insecurity is predicted by macroeconomic variables such as actual GDP, output gap, inflation, unemployment rate, investment, household consumption and saving, and so on. As demonstrated by the literature on Dynamic Stochastic General Equilibrium (DSGE) models, these variables move along one or multiple equilibrium paths in response to exogenous shocks that are more or less persistent and potentially capable of triggering phases of expansion and recession (Dave and Sorge, 2025). Governments and central banks react to cyclical fluctuations by using the economic policy tools at their disposal (taxation, public spending, social transfers, interest rates, money supply) to stabilize price and national income growth, preserve employment, and contain the harmful effects of adverse shocks. This macroeconomic dynamics inevitably reflects on the expectations of households and firms about the fundamentals and future perspectives of being unemployed. The lack of studies focused on the macroeconomic dimension does not allow for discerning the short-run and long-run effects of exogenous shocks on job insecurity, nor for defining public policies to combat job insecurity that adequately account for the effects of the business cycle. In fact, to effectively design and implement their policies for containing job insecurity, policymakers need analytical tools that consider the gradual processes of divergence and convergence of the actual unemployment rate with respect to its steady-state equilibrium (which, in turn, are strictly dependent on cyclical fluctuations). This manuscript attempts to fill this gap by proposing a rational expectations New Keynesian model populated by public agents (central bank and government) who are fully informed and private agents (firms and households) who directly observe only a part of the state variables and who receive a potentially incomplete and noisy informative signal regarding unobservables and informative shocks (which is used by them to formulate their long-run expectations about the fundamentals). The expression “potentially incomplete and noisy” indicates that the information transmitted by public agents to private ones can be partial (i.e., not include all values of the unobservables) and flawed by ambiguity. The main innovation of the work consists in demonstrating that adopting a fully transparent communication strategy represents a sufficient condition to guarantee equilibrium determinacy and uniqueness, regardless of whether the Taylor Principle is satisfied (i.e., the central bank responds aggressively to an increase in prices by raising its policy rate more than proportionally to the inflation rate). However, along this equilibrium path, the “Paradox of Transparency'” can emerge, meaning that the transmission of otherwise unobservable information by public agents to private ones can increase job insecurity and lead to a loss in terms of welfare (Morris and Shin, 2005). The predictions of the theoretical model are validated by the multilevel procedure proposed by Lippi (2021), which uses Principal Component Analysis to compute the exogenous shocks and the High-Dimensional Dynamic Factor model to estimate the dynamic factor loadings associated with the endogenous variables along their balanced growth paths. This technique is applied to time-series data ranging from the second quarter of 1999 to the first quarter of 2025 related to three large economies of the Mediterranean, namely, France, Italy, and Spain. The results of the empirical analysis point out that the lack of institutional transparency is responsible for the emergence of the higher-order moments displayed by the observed data (skewness and excess of kurtosis) and confirm that the Paradox of transparency can emerge even in the common knowledge setting. Although the current model doesn’t explicitly model age heterogeneity, future research will validate a calibrated New Keynesian model on aggregate and mature-worker data to estimate job insecurity. Subsequently, we will compare the theoretical series (under common knowledge) with the proxy from the SHARE data in Europe /2004–2022), which reflects the (equilibrium) effects of non-transparent communication. This vis-à-vis enables a counterfactual on the impact of full institutional transparency for older workers.
Job insecurity, ageing and institutional transparency: Theory and time-series evidence from the Euromed area
Luca Vota
Primo
2025
Abstract
Precarious work is a widespread phenomenon in both advanced economies and in emerging and developing countries and remains a major policy concern in many contexts. According to a survey conducted by the Ipsos institute on 12,000 adult workers residing in 27 countries (Ipsos 2020), 54% of the sample were concerned about losing their job in the 12 months following the interview (17% “very concerned”, 37% “fairly concerned”). The highest values were recorded in Russia (75%), Spain (73%), and Malaysia (71%), while the lowest ones were reported in Germany (26%), Sweden (30%), the Netherlands, and the United States (36%). Although they indicate strong variability between countries, data demonstrate that job insecurity, defined as the subjectively perceived probability of experiencing an interruption in one's career (Shoss, 2017), is a pervasive issue even in nations with a strong industrial vocation and generous welfare measures (as in the case of Germany and Swede Since the 1990s, global structural shifts loosened traditional worker–firm ties (Davis, 2013).This sparked sustained scholarly interest in job insecurity. Globalization and automation cut costs and cycle times, yet heightened perceived insecurity (Scheve & Slaughter, 2006; Couple, 2019; Nam, 2019; Cao & Song, 2025). Technological progress threatens not only low-skilled workers, but also highly qualified professionals (Colvin, 2015). The diffusion of Artificial Intelligence has amplified these concerns, with similar effects for managerial and non-managerial staff (Koo et al., 2021). AI’s negative impact is partly mediated by vocational learning ability - the capacity to autonomously acquire and apply new skills (Liu & Zhan, 2020). Constant AI interaction pushes continuous upskilling, shaping insecurity, creativity, and well-being (Wu et al., 2024). Fear of rapid skill obsolescence - and of human roles being replaced by virtual ones - adds further stress (Sharif et al., 2025). The approach adopted by researchers to date has favored the analysis of individual and organizational drivers of job insecurity, paying little attention to systemic ones. In particular, they have used survey data relating to variables such as contract type, responsibilities associated with one's position, hourly wage, health conditions, number of children, age, years of experience, HR policies and so on. Such analyses have often focused on specific sectors and regions to account for the heterogeneity present in different economic sectors and in the economic conditions of various countries, proving useful in evaluating the predictors of job insecurity in different contexts (Lee et al., 2018; Chirumbolo et al., 2020; Martínez et al., 2020; De Cuyper et al., 2021; Muñoz Medina et al., 2023; Darvishmotevali, 2025). However, a fundamental limitation of the individual-based approach is its inability to account for macroeconomic variables and (consequently) the influence that the alternation of the various phases of the business cycle exerts on job insecurity. Indeed, while purely individual variables undoubtedly have a significant impact on job insecurity, it is equally plausible that a non-negligible component of insecurity is predicted by macroeconomic variables such as actual GDP, output gap, inflation, unemployment rate, investment, household consumption and saving, and so on. As demonstrated by the literature on Dynamic Stochastic General Equilibrium (DSGE) models, these variables move along one or multiple equilibrium paths in response to exogenous shocks that are more or less persistent and potentially capable of triggering phases of expansion and recession (Dave and Sorge, 2025). Governments and central banks react to cyclical fluctuations by using the economic policy tools at their disposal (taxation, public spending, social transfers, interest rates, money supply) to stabilize price and national income growth, preserve employment, and contain the harmful effects of adverse shocks. This macroeconomic dynamics inevitably reflects on the expectations of households and firms about the fundamentals and future perspectives of being unemployed. The lack of studies focused on the macroeconomic dimension does not allow for discerning the short-run and long-run effects of exogenous shocks on job insecurity, nor for defining public policies to combat job insecurity that adequately account for the effects of the business cycle. In fact, to effectively design and implement their policies for containing job insecurity, policymakers need analytical tools that consider the gradual processes of divergence and convergence of the actual unemployment rate with respect to its steady-state equilibrium (which, in turn, are strictly dependent on cyclical fluctuations). This manuscript attempts to fill this gap by proposing a rational expectations New Keynesian model populated by public agents (central bank and government) who are fully informed and private agents (firms and households) who directly observe only a part of the state variables and who receive a potentially incomplete and noisy informative signal regarding unobservables and informative shocks (which is used by them to formulate their long-run expectations about the fundamentals). The expression “potentially incomplete and noisy” indicates that the information transmitted by public agents to private ones can be partial (i.e., not include all values of the unobservables) and flawed by ambiguity. The main innovation of the work consists in demonstrating that adopting a fully transparent communication strategy represents a sufficient condition to guarantee equilibrium determinacy and uniqueness, regardless of whether the Taylor Principle is satisfied (i.e., the central bank responds aggressively to an increase in prices by raising its policy rate more than proportionally to the inflation rate). However, along this equilibrium path, the “Paradox of Transparency'” can emerge, meaning that the transmission of otherwise unobservable information by public agents to private ones can increase job insecurity and lead to a loss in terms of welfare (Morris and Shin, 2005). The predictions of the theoretical model are validated by the multilevel procedure proposed by Lippi (2021), which uses Principal Component Analysis to compute the exogenous shocks and the High-Dimensional Dynamic Factor model to estimate the dynamic factor loadings associated with the endogenous variables along their balanced growth paths. This technique is applied to time-series data ranging from the second quarter of 1999 to the first quarter of 2025 related to three large economies of the Mediterranean, namely, France, Italy, and Spain. The results of the empirical analysis point out that the lack of institutional transparency is responsible for the emergence of the higher-order moments displayed by the observed data (skewness and excess of kurtosis) and confirm that the Paradox of transparency can emerge even in the common knowledge setting. Although the current model doesn’t explicitly model age heterogeneity, future research will validate a calibrated New Keynesian model on aggregate and mature-worker data to estimate job insecurity. Subsequently, we will compare the theoretical series (under common knowledge) with the proxy from the SHARE data in Europe /2004–2022), which reflects the (equilibrium) effects of non-transparent communication. This vis-à-vis enables a counterfactual on the impact of full institutional transparency for older workers.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
