Working Paper
Paradisi M, Bovini G. Labor Substitutability and The Impact of Raising the Retirement Age [JOB MARKET PAPER]. Working Paper.Abstract
The degree of substitutability among different types of workers can affect the incidence of public policies by creating labor demand spillovers within firms. We study the substitutability between age cohorts in the context of an unexpected public pension reform increasing the full retirement age in Italy starting in 2012.  An increase of the full retirement age creates incentives for older workers to delay retirement and temporarily increases their retention at the employer firm. We investigate how retaining an additional senior worker affects employer labor demand for other workers. Using a large and rich matched employer-employee administrative data set for Italy, we exploit the substantial idiosyncratic firm-level variation in the impact of the unanticipated pension reform even after conditioning on the broad demographics of firms' work forces. We find that workers on the cusp of retirement and younger co-workers are substitutes. When a workers is retained firms fire on average 0.16 workers more and decrease hiring by 0.29 units. Older workers in the same occupation group are the closest substitute to the retained older workers who delay retirement. Extending the retirement age also decreases labor earnings and increases the take-up of social security programs for incumbent workers. We evaluate the implications of the observed patterns of labor-labor substitutability for the incidence of the policy by estimating the fiscal cost of the public pension reform in the short-run. We use a government accounting model that allows for both substitution across government programs and for spillovers from delayed retirement onto other incumbent workers. Our findings indicate that the labor demand spillovers explain almost all the short-run fiscal cost of the policy in the first four years after implementation.
Bianchi N, Bovini G, Li J, Paradisi M, Powell M. Chains of Opportunity Revisited. Working Paper.Abstract
Promotions are an important component of a worker’s wage. Yet, traditional theories about the factors driving career progressions typically focus on worker-level characteristics like human capital acquisition, on learning, or on broad market-level factors like labor supply and demand. We study coworker-career spillovers that arise in firms with limited promotion opportunities. We exploit a 2011 Italian pension reform that tightened eligibility criteria for the public pension. We use administrative data on Italian private-sector and leverage cross-firm variation to isolate the effect of retirement delays among soon-to-retire workers on the promotions of their colleagues. We find evidence of career spillovers, and the patterns of these spillovers are consistent with the idea that older workers block the careers of their younger colleagues in firms with limited opportunities. Delays in retirement lead to a decrease in younger workers’ wage growth. Promotions from blue to white-collar positions fall in response to retirement delays among white-collar workers, whereas there is no effect of such delays among blue-collars. The effects are largest in firms with shrinking employment in the years leading up to the policy and negligible among fast-growing firms. We derive in a model the key features necessary to explain our results.  
In Preparation
Paradisi M. Firm-level Tax Incidence: the Redistribution of the Cost of Labor. In Preparation.Abstract

How is the cost of payroll taxes redistributed within the firm? Standard tax incidence theory is based on the assumption that labor market are perfectly competitive and firm-level shocks to the cost of labor have no effects on the wages of incumbent workers. Identification techniques applied at the individual level balance unobservable individual characteristics and by construction rule out firm idiosyncratic characteristics. As a result, the estimated elasticity misses an important channel of the pass-through of taxes to wages. I attempt to identify this component by studying how wages respond to firm-level changes in the cost of labor. I use administrative data for the universe of French private sector firms that I match to information on paid social security contributions to study a payroll tax reform that was implemented in France in 2003. The policy changed payroll tax subsidies for low-paid workers leaving tax rates unchanged for other workers. I exploit variation in the wage distribution of the subset of low-paid workers to create firm-level shifts in the average payroll tax rate of low-paid employees and I use them as an instrument for the total change in the cost of labor. I study spillovers within the firm by looking at the workers who are unaffected by the policy and I evaluate how an extra euro of subsidies to low-paid individuals is redistributed along the wage distribution. Preliminary results show that an increase in the average tax rate of low-paid employees causes significant drops in the wages of incumbent high-paid employees.

Gregorio ED, Paradisi M. In the Business of Compliance: Firm Revenues Manipulation in Response to the Audit System. In Preparation.Abstract

Small and medium businesses face greater opportunities to conceal their income sources relative to employees, leading to substantial tax gaps. Despite the magnitude of this problem, evidence on how these businesses comply with the tax system is still scarce. We exploit a unique discontinuity in audit probability around a firm-specific revenue threshold in the Italian audit system to estimate revenue manipulation elasticities and perceived audit risk. To this end, we rely on new sources of administrative data on the universe of tax returns filed by small businesses and self-employed over 2007-2016 in Italy. Our theory adapts non-linear budget set models to the context of firm revenue declaration. Unlike previous studies using these methods, the probability jump in our setting allows to directly identify the elasticity of revenue manipulation as a component of the total taxable revenue elasticity. Our framework also provides us a non-parametric upper bound to the share of non-manipulators, which we use to characterize the correlates of business tax compliance.