Research

"For today’s applied economists, parallelism has become the default lens on reality."

"Valid frequentist causal inferences from an observational study can only be obtained if you embed the observational study within a hypothetical randomized experiment... otherwise you are just looking at computer output and implied associations, and making up, generally silly, although sometimes interesting, stories" (Don Rubin)

"And the less data we have, the more assumptions we have to make, the more predictions become narratives." (Christopher Molnar)

Personal Research Links: Ideas ResearchGate Google Scholar

Journal Publications

On the (Mis)Use of Machine Learning with Panel Data (with M. Letta, G. Pinto) Oxford Bulletin of Economics and Statistics, published online open access, 12 September 2025.

Abstract: We provide the first systematic assessment of data leakage issues in the use of machine learning on panel data. Our organising framework clarifies why neglecting the cross-sectional and longitudinal structure of these data leads to hard-to-detect data leakage, inflated out-of-sample performance, and an inadvertent overestimation of the real-world usefulness and applicability of machine learning models. We then offer empirical guidelines for practitioners to ensure the correct implementation of supervised machine learning in panel data environments. An empirical application, using data from over 3000 U.S. counties spanning 2000 to 2019 and focused on income prediction, illustrates the practical relevance of these points across nearly 500 models for both classification and regression tasks.

The Bright Side of Breaking Up: How Territorial Fragmentation Increases Political Engagement in Local Communities (with Di Matteo, D.) Journal of Regional Science, published online 10/04/2025. Media coverage: Senato della Repubblica

Abstract: Territorial reforms of administrative boundaries primarily aim at pursuing cost and administrative efficiency objectives, but their impact on the political engagement of local communities remains unclear. Moreover, while amalgamations have been widely studied, little is known about the effects of territorial fragmentation. To bridge this gap, we examine a regional reform in Italy's Apulia region, wherein five municipalities voluntarily underwent division in the mid-1970s. Through examining its long-term effects on voter turnout, we uncover new insights on the potential merits of territorial fragmentation in fostering enhanced political engagement, especially at the local level.

The Municipal Administration Quality Index: The Italian case (with C. Giannantoni, M. Mazziotta, F. Zampollo) Social Indicators Research, published online 21/01/2025. WP version. Companion website

Abstract: Municipal governments have become increasingly involved in public service provision, fund management, and administrative tasks. Despite their importance, studies of their relevance for local economic development have lacked a measure of their quality and capacity with extensive temporal and territorial coverage. We introduce the Municipal Administration Quality Index (MAQI), a novel composite index designed to assess the quality of local governments in Italy. Drawing on a newly compiled dataset covering 2001–2022, MAQI evaluates three key pillars: bureaucratic efficiency and capacity, the quality of local politicians, and the economic and fiscal performance of municipal administrations. Our findings reveal significant variations in administrative quality, with larger municipalities exhibiting higher scores due to fewer financial constraints and better-educated civil servants. We also identify distinct geographical patterns in institutional quality, challenging the conventional North–South divide.

Do green parties affect local waste management policies? (with N. Fiorino, E. Galli) Journal of Environmental Economics and Management, 128, 103056, 2024. WP version. Media coverage: LaVoce

Abstract: We explore whether mayors supported by pro-environmental parties enhance local environmental outcomes compared to their non-environmental counterparts. We study close elections within a regression discontinuity design and find a notable rise in recycling rates in Italian municipalities governed by pro-environmental coalitions. This uptick becomes far less pronounced when adopting broader criteria to define green mayoral candidates. Crucially, the enhanced recycling rates are not realized through augmented budgets for environmental initiatives or waste collection, but rather are primarily attributed to the implementation of local policies, such as on-call waste collection and the establishment of waste collection centers.

A decade of Eurasian Integration: An ex-post non-parametric assessment of the Eurasian Economic Union (with P., Montalbano, Z., Temerbulatova) International Economics, 178, 100506, 2024.

Abstract: This paper provides a sound ex-post evaluation of the impact of the Eurasian integration on member countries’ bilateral trade after a decade of implementation. We overcome the main limitations of current empirical analyses on the effects of trade agreements, namely the aggregation of tariff and non-tariff barriers and the likely self-selection bias, by applying a non-parametric method specifically designed to fully exploit time-series cross-sectional data. We thus compare the trade flows of the member countries in the Eurasian agreement with the exporter-importer pairs located in the Eurasian continent, which are most similar in terms of pre-treatment trends and features. Our results confirm the previous literature about the lack of a significant impact of the Eurasian customs union but find more positive net effects of the more recent integration steps. Our results ask for additional efforts to complete the Eurasian integration and let its member countries fully benefit from its hoped-for long-term effects.

The long-term effects of mass layoffs: do local economies (ever) recover? (with V., Celli, G., Pellegrini) Journal of Economic Geography, 23, 1121-1144, 2023. WP version

Abstract: This article investigates the long-term reaction of local labor markets (LLMs) to a mass layoff in a manufacturing plant. We adopt a non-parametric generalization of the difference-in-differences estimator expressly developed for time-series cross-sectional data and a new comprehensive dataset. Our results suggest that, on average, a mass layoff abruptly decreases industry employment by 22%; this negative impact is persistent even 9 years later. The shock has a negative and statistically significant effect only on the same industry of the affected LLM, while the rest of the local economy is, at most, mildly negatively affected. These findings depend on the initial level of development and call for the policymakers’ intervention to design efficient employment policies aimed at reducing the long-lasting social costs of a mass layoff at least for the less developed and less dynamic local economies.

Electoral earthquake: local shocks and authoritarian voting (with C., Ferrante, M., Letta) European Economic Review, 156, 104464, 2023. OPEN ACCESS Media coverage: VoxEU column, Nacio. WP version

Abstract: A growing literature has highlighted the role of economic grievances, global transformations, cultural cleavages and long-term trends of isolation and decline in engendering political discontent. However, this literature is silent on the potential role of unanticipated local shocks in fuelling support for authoritarian parties. We fill this gap by using comprehensive data at a fine spatial scale and a comparative natural experiment approach. Our study documents that the occurrence of two destructive earthquakes in Italy resulted in sharply diverging electoral outcomes: while the 2012 Emilia quake did not alter voting behaviour, the 2009 L’Aquila earthquake paved the way for an impressive and persistent authoritarian backlash in the most affected areas. Such heterogeneous patterns originate from a stark contrast in post-disaster reconstruction processes and shifts in institutional trust. These findings suggest that valence issues generated from local shocks can turn “places that don’t recover” into authoritarian hotbeds.

What kind of region reaps the benefits of a currency union? Journal of Regional Science, 63(3), 552-582, 2023. (with R. Di Stefano, G. Pellegrini).

Abstract: What is the economic impact of joining a currency union? Is this impact heterogeneous across regions? And how does it change in case of a recession? We answer these questions by investigating the economic impact of joining the euro area for the latecomers, i.e., the eastern European countries that adopted the euro after 2002. Differently from previous literature, we use NUTS-2 regions as units of analysis. This novelty allows us to investigate the theoretical predictions of a currency union impact at a more appropriate geographical level. Using a recently developed counterfactual approach, we estimate the overall as well as the disaggregated impact of joining the euro area. We find that the adoption of the euro brought about a small positive effect, which was, however, dampened by the Great Recession. Individual regional estimates suggest heterogeneous returns with benefits accruing mostly to core regions.

Deeds or words? The local influence of anti-immigrant parties on foreigners’ flows. European Journal of Political Economy, 77, 102275, 2023. (with F. Zampollo). Media coverage: LaVoce; OpenPolis; la Repubblica

Abstract: We investigate the influence of anti-immigrant parties on foreigners' location choices. Considering Italian municipal elections from 2000 to 2018, we create a comprehensive database that includes a classification of the anti-/pro-immigration axis of leading political parties based on specialists' assessments. Adopting a bias-corrected regression discontinuity design, we find that the election of a mayor supported by an anti-immigrant coalition significantly affects immigrants' location choices only when considering the most recent years. This finding is not driven by the enactment of policies against immigrants but by an ‘inhospitality effect’, which has become stronger over time due to the exacerbation of political propaganda. Therefore, foreigners' flows are influenced by the local political environment only when immigration is central to the political debate.

I will survive! The impact of place-based policies when public transfers fade out. Regional Studies, 2022. (with G. Pellegrini).

Abstract: Are place-based policies capable of taking lagging areas to a higher growth trajectory permanently? We answer this crucial question by investigating what happens when strongly subsidised regions experience a substantial reduction in funding. By analysing an extensive database via the mean balancing approach, we estimate the causal impact of exiting the convergence region status of the EU regional policy. We find that only regions which experienced a considerable reduction in funding during a recession suffered a negative impact on economic growth. However, the impact varies with the features of the regions and the local economic context.

Can regional policies shape migration flows? Papers in Regional Science, 101(3), 515-536, 2022. OPEN ACCESS (with G. Pellegrini, O. Tarola). Poster; WP version

Abstract: Our empirical analysis focuses on the effect of regional policies on migration attraction factors in Europe. We employ a regression discontinuity design to assess the causal relationship between the reception of large amounts of public funds and migration flows in the EU-15 regions. In highly-subsidized regions, we find a large increase in the share of foreign citizens from less-developed countries when compared to low-subsidized regions with similar pre-treatment characteristics. The analysis shows that such an increase is due to the positive impact of the European regional policy on job market opportunities as well as the improvement of public goods supply.

Local inequalities of the COVID-19 crisis. Regional Science and Urban Economics, 92, 103752, 2022. (with M., Letta). Media coverage: VoxEU column; LaVoce; Il Venerdì di Repubblica; icalabresi.it

Abstract: This paper assesses the pandemic’s impact on Italian local economies with the newly developed machine learning control method for counterfactual building. Our results document that the economic effects of the COVID-19 shock vary dramatically across the Italian territory and are spatially uncorrelated with the epidemiological pattern of the first wave. The largest employment losses occurred in areas characterized by high exposure to social aggregation risks and pre-existing labor market fragilities. Lastly, we show that the hotspots of the COVID-19 crisis do not overlap with those of the Great Recession. These findings call for a place-based policy response to address the uneven economic geography of the pandemic.

Decomposing the employment effects of investment subsidies. Journal of Urban Economics, 128, 103408, 2022. (with G., Pellegrini).

Abstract: Most governments tackle the economic issues of underdeveloped areas by offering subsidies aimed at fostering economic activities and local employment. Localized policies put constraints on where businesses may locate to receive subsidies, but they generally place few restrictions on whom subsidized businesses must hire. Using administrative data on firms and workers in Italy, we adopt a multi-cutoff regression discontinuity design to empirically assess and decompose the employment effect of substantial incentives for the replacement or establishment of new capital. Our empirical strategy allows identifying the geographical origin and labor market status of new hires. The results show how the majority of recruits come from new entrants to the labor market, in particular, young people and students, while displacement effects are limited. It appears that subsidized companies tend to keep their most valuable staff and hire more qualified young people. Overall, we find only a modest spatial dispersion of the effects or a possible crowding-out of the local labor market.

How much does state aid mitigate employment losses? Local policy effects at a time of economic crisis. Regional Studies, 56(10), 1698-1712, 2022. (with G., Pellegrini).

Abstract: We evaluate the effectiveness of the most extensive experiment of income redistribution, i.e., the EU regional policy, at a time of economic crisis. By exploiting geographic discontinuities in fund eligibility, we analyze comprehensive data on all publicly funded Italian projects at the municipality level. We find a positive impact of localized support to firms with a sizable increase in employment and the number of plants in intensively treated areas. The result is an important policy lesson concerning the effects of place-based policies at a time of a long-lasting recession, such as that engendered by the COVID-19 crisis.

Does R&D expenditure boost economic growth in lagging regions? Social Indicators Research, 2021, published online: 25 September 2021. OPEN ACCESS (with V., Celli, G., Pellegrini).

Abstract: We assess the impact of the EU Regional Policy on regional economic growth by applying a new evaluation strategy, which integrates mediation analysis with a quasi-experimental framework. Using the R&D expenditure as an indicator of innovation capability, we evaluate how much of the total effect of the EU Regional Policy is due to R&D in the poorest EU regions. Consistently with the previous literature, we found a positive impact of the overall policy on economic growth, but, among the convergence regions, those investing a higher proportion of funds in R&D have the same convergence rate as regions investing more in other priorities. These findings confirm that the EU Regional Policy played an important role in the economic recovery of the poorest regions in the aftermath of the Great Recession. However, focusing resources on R&D does not seem to provide additional economic benefits, at least in the short run.

Local mortality estimates during the COVID-19 pandemic in Italy. Journal of Population Economics, 34, 1189–1217, 2021. OPEN ACCESS (with Roberta Di Stefano, Marco Letta, Sara Miccoli). WP version

Abstract: Estimates of the real death toll of the COVID-19 pandemic have proven to be problematic in many countries, Italy being no exception. Mortality estimates at the local level are even more uncertain as they require stringent conditions, such as granularity and accuracy of the data at hand, which are rarely met. The ‘official’ approach adopted by public institutions to estimate the ‘excess mortality’ during the pandemic draws on a comparison between observed all-cause mortality data for 2020 and averages of mortality figures in the past years for the same period. In this paper, we apply the recently developed machine learning control method to build a more realistic counterfactual scenario of mortality in the absence of COVID-19. We demonstrate that supervised machine learning techniques outperform the official method by substantially improving the prediction accuracy of local mortality in ‘ordinary’ years, especially in small- and medium-sized municipalities. We then apply the best-performing algorithms to derive estimates of local excess mortality for the period between February and September 2020. Such estimates allow us to provide insights about the demographic evolution of the first wave of the pandemic throughout the country. To help improve diagnostic and monitoring efforts, our dataset is freely available to the research community.

When did coronavirus arrive in Europe? Statistical Methods and Applications, 31, 181–195, 2022. OPEN ACCESS (with Roberta Di Stefano). WP version

Abstract: The first cluster of coronavirus cases in Europe was officially detected on 21st February 2020 in Northern Italy, even if recent evidence showed sporadic first cases in Europe since the end of 2019. In this study, we have tested the presence of coronavirus in Italy and, even more importantly, we have assessed whether the virus had already spread sooner than 21st February. We use a counterfactual approach and certified daily data on the number of deaths (deaths from any cause, not only related to coronavirus) at the municipality level. Our estimates confirm that coronavirus began spreading in Northern Italy in mid-January.

Evaluation of Education and Training Impacts for the Unemployed: Challenges of New Data. Labour Economics, 67, 1-20, 2020. (with Urwin, P., Bibby, D., Thomson, D.)

Abstract: This study utilises an exceptionally rich English administrative dataset, to estimate employment impacts from training voluntarily initiated by unemployed individuals. A Coarsened Exact Matching approach is adopted, in a dynamic evaluation framework, to estimate impacts up to 5 years from training start. We identify economically and statistically significant impacts, estimated separately for (i) all training starters, (ii) the partially, and (iii) fully treated. Investigation of possible endogenous selection into partial/full treatment, using distance to training provider as an instrumental variable, suggests inclusion of extensive employment and learning histories in a matching framework, justifies invocation of the conditional independence assumption for comparisons of full/partial treatment. The partially treated secure a return that is, on average, 2 percentage points lower than full treatment. Thus, an ‘intention to treat’ approach would not alter conclusions on the efficacy of training; but using the partially treated to estimate counterfactual outcomes risks understating returns.

Local multipliers at work. Industrial and Corporate Change, 29(4), 959–977, 2020. (with Guido Pellegrini) Poster (Working paper version: Local multipliers at work. Earlier version in Italian: Qual è l'impatto delle politiche regionali sul mercato del lavoro?).

Abstract: We propose a novel evaluation strategy to estimate local multipliers in Italy during 1996-2006. We find the presence of positive multipliers: 0.26-0.33 for the tradable sector (manufacturing) and 0.88-1.13 for the non-tradable sector (construction and services). They are lower than what was previously found for the US but much higher than those identified for European and Asian countries. The reasons for this finding lie in the higher accuracy of the data, in the relevance of the instrument used, and in the widespread underutilization of production factors.

Endogenous amenities, tourists’ happiness and competitiveness. Regional Studies, 54(9), 1214-1225, 2020. (with Cristina Bernini, Guido Pellegrini) (Working paper version with previous title Place-based amenities, well-being and territorial competitiveness: a new approach using tourists’ happiness)

Abstract: A key strategy for supporting destination competitiveness is to enhance endogenous amenities, and tourists are the best candidate to evaluate them at the destination. The analysis in this paper uses a comprehensive data set on foreign travellers to investigate their happiness at Italian destinations between 2005 and 2014. Using a theory-dependent approach to model happiness at the destination with respect to endogenous and exogenous amenities, personal characteristics and trip features, a great diversity in the mix of amenities affecting tourist happiness is shown. However, some clear spatial patterns emerge. The findings call for place-based policies targeted at the specific needs of each area.

Are eco-labels good for the local economy? Papers in Regional Science, 99(3), 645-661, 2020. (with Cristina Bernini) OPEN ACCESS (Working paper version with previous title Do sustainability policies finance local economies?)

Abstract: We evaluate whether the adoption of a well-known transition management instrument in the tourism industry can support simultaneously economic growth as well as sustainability. We create a detailed dataset at the municipality level and use a recently developed policy evaluation technique to investigate the causal impact of the Blue Flag programme on the local economies. Estimates show that this eco-label is not effective at enhancing the local economy; findings are homogeneous across destinations. This empirical result is in line with the recent theoretical literature arguing that a single policy does not suffice for transiting towards a sustainable economy.

Are we spending too much to grow? The case of Structural Funds. Journal of Regional Science 58(3), 535-563, 2018. (with Guido Pellegrini). Press coverage: 1 2 3

Abstract: We evaluate whether the impact of EU Structural and Cohesion Funds (EUF) on Member States’ regional economic growth depends on the intensity of treatment, measured by the normalized amount of funds distributed in each region. We use an original data set that covers all the main sources of EUF and extend the regression discontinuity design to the case of continuous treatment. The results suggest an average positive effect on regional growth. The estimated conditional intensity-growth function is concave and presents a maximum value. Therefore, the exceeding funds could have been allocated to other lagging regions without reducing the effect on growth.

Industrial policy evaluation in the presence of spillovers. Small Business Economics 49(3), 671-686, 2017. OPEN ACCESS (with Guido Pellegrini) (Working paper version with previous title Beyond the SUTVA: how policy evaluations change when we allow for interactions among firms).

Abstract: The shortage of studies on spatial spillovers of capital subsidy policies is rather surprising, considering that such policies are usually designed to generate spatial externalities. We propose a new framework that allows positive agglomeration effects to be contrasted with the negative cross-sectional substitution and the crowding-out effect. The global evaluation of the ATT and the spillover parameters shifts the spotlight from the policy effect on subsidised firms to the global effect of capital subsidy policies on the targeted territory. The empirical evaluation of a policy in Italy mainly directed towards small- and medium-sized firms shows that the impact on investments, turnover and employment is positive and large, but is negative on TFP. However, the employment growth is partially determined to the detriment of the untreated firms.

Public subsidies, TFP and efficiency: a tale of complex relationships. Research Policy 46(4), 751–767, 2017. (with Cristina Bernini, Guido Pellegrini) Working paper version

Abstract: This paper evaluates the impact of subsidies on the different components of TFP for granted firms’ longterm growth. The impact of capital subsidies is captured by a quasi–experimental method (Multiple RDD), exploiting the conditions for a local random experiment created by an Italian industrial policy. Results show that capital subsidies negatively affect TFP growth in the short term, and signals of positive effects appear only after 3–4 years. This positive medium-long term impact comes especially through technological change and not through scale impact change, as may have been expected.

Do Subsidies to Private Capital Boost Firms’ Growth? A Multiple Regression Discontinuity Design Approach. Journal of Public Economics 109, 114-126, 2014. (with Guido Pellegrini)

Abstract: There is still little consensus among economists on the effectiveness of business support policies. The evaluation of such policies requires a reliable identification procedure that is hardly achieved in empirical studies. We analyse the impact of a policy instrument – Law 488/92 (L488), the main Italian regional policy – that allocates subsidies to private firms by a multiple ranking system. Thanks to the peculiar L488 selection process that creates the conditions for a local random experiment, we are able to assess the effectiveness of these types of incentives for a relevant subgroup of firms. We propose a nonparametric multiple rankings regression discontinuity design that exploits the sharp discontinuities in the L488 rankings and extends the regression discontinuity design (RDD) approach to a context where the treatment is assigned by multiple rankings with different cut-off points. We find that the impact of the subsidies on employment, investment, and turnover is positive and statistically significant, while the effect on productivity is mostly negligible. The new subsidised capital is additional but non-complementary with the owner-financed investment. The results are robust to different specifications and not due to intertemporal substitution.

(Other) Journal Publications

The Italian Geography of Discontent. Scienze Regionali, 3, 367-384, 2022. (with M., Letta, F. Zampollo).

Abstract: Although widespread, the recent populist wave in Western countries is a heterogeneous phenomenon in terms of individual features of populist voters — the stereotype is “older, working-class, white, poorly educated, who live on low incomes” — as well as geographical characteristics of populist hotspots — “lagging-behind, stagnating and low-productivity regions”. This study leverages nonlinear statistical learning techniques to detect recurrent individual and geographical patterns of populist voting across Italy. Using the Chapel Hill expert survey classification, we analyse the most prominent voting patterns during the 2019 European elections in all Italian local labour markets. We map the Italian geography of discontent, highlighting how it seems to be shaped by the interaction between individual- and territorial-level predictors. Our study promotes the adoption of flexible and nonparametric predictive algorithms to ‘diagnose’ the main factors linked to the spatial distribution and evolution of populist hotspots.

Evaluation of the Effectiveness of Firm Subsidies in Lagging-Behind Areas: The Italian Job. Scienze Regionali, 3/2020, pp. 477-500, 2020. (with Guido Pellegrini)

Abstract: Since the late 1990s, Italian scholars have produced numerous studies in the field of regional policy evaluation, especially ones that have investigated the impact of financial incentives aimed at supporting the accumulation of private capital in underdeveloped areas. The number and innovativeness of these studies make it possible to define the presence of an Italian school for evaluating regional policies. This paper testifies to the importance and methodological advances of this school, putting it at the frontier of policy evaluation analyses. The presentation of the studies moves in two directions, historical and methodological, identifying the main themes and techniques addressed in recent years: the evaluation of Law 488 and negotiated programming policies, on the one hand, the advance in policy evaluation techniques in the presence of interactions and continuous treatment, on the other. The paper does not claim to be an exhaustive review; rather, it should be considered an overview of the historical path and the future prospects of what we call ‘the Italian school of regional policy evaluation’.

Quality competition and environmental damage: is there a role for relative preferences? Journal of Cleaner Production, 266, 1-13, 2020. (with Guido Pellegrini and Ornella Tarola)

Abstract: We consider from a theoretical and empirical viewpoint how network preferences of consumers, which differ in their willingness to pay (WTP) for green goods, affect the equilibrium configuration of a vertically differentiated duopoly when a cleaner firm competes against a dirtier rival. We find that a trade-off can emerge between emissions abatement and the catching-up process of the dirtier firm can emerge. In particular, emissions decrease only when consumers with low WTP have network preferences. Still, in this case, only the cleaner firm has an incentive to improve its environmental quality along the quality ladder, while the laggard firm is tied to its backward profile. These theoretical predictions are empirically tested by exploiting the duality of Italy's economy, with high-income citizens mainly concentrated in Northern regions, while low-income citizens are largely located in the South. We adopt a fixed-effects spatial error model at the provincial level to relate variables on firms' environmental management system with citizens' environmental behaviour, proxied by the recycling rate. Overall, the empirical estimates are in line with the theoretical predictions. From the empirical analysis, it emerges that consumers' environmental behaviour positively affects enterprises' future environmental choices. Moreover, firms have a stronger incentive to raise their environmental quality when low-income consumers, namely those buying dirtier variants, internalise the network effect.

Quantitative Evaluation Techniques for Regional Policies. In R Capello, P Nijkamp (ed.) Handbook of Regional Growth and Development Theories, Second Edition, Edwar Elgar Publishing, chapter 28, 2019. (with Guido Pellegrini)

Abstract: In recent years, the use of counterfactual techniques for the evaluation of regional policies has greatly expanded, mainly due to the availability of ever better data from the point of view of geographical location, the characteristics of the subjects involved, the temporal and spatial coherence. This process has posed new challenges to the econometric techniques developed to estimate regional policy impacts. There are specific characteristics of place-based policies that necessarily require a particular specification or even a specific adaptation of counterfactual techniques. The greatest difficulty is in the inherent endogeneity of place-based policies: the lower the development of a region, the greater the public intervention. In addition, the presence of interferences between treated subjects, between untreated subjects and between both, leads to the need to adapt the Rubin causal model, as it is based on the Stable Unit Treatment Value Assumption, which postulates that there is no interference between units. Furthermore, the presence of spillovers due to interference creates the need to define various measures of policy impact, considering direct, indirect, and total effects. This survey gives an account of the evaluation techniues most used in such literature, and of some recent or still evolving methodological developments.

Intensità ed efficacia dei Fondi Strutturali Europei. EyesReg 7(6), November 2017. (with Guido Pellegrini)

The signalling effect of eco-labels in modern coastal tourism. Journal of Sustainable Tourism 8, 1159-1180, 2017.

Abstract: As the demand for environmentally sustainable tourism grows, eco-labels are becoming increasingly popular as a signal of environmental quality. However, the existence of a causal link between awarding a seaside eco-label and the increase in tourism flows is still under discussion in the literature. In this article, we gauge the signalling impact of a specific eco-label, the Blue Flag award, using detailed data on tourism flows to seaside Italian destinations during the period 2008–2012. We adopt a recent econometric modelling strategy – the synthetic control method – in shaping estimation results and testing the sensitivity and robustness of our results. We find that being awarded the Blue Flag increases the flow of domestic tourists for up to three seasons after assignment. However, we find no effect for the flow of international tourists. Investigating the mechanisms driving the results, we find that the award of a Blue Flag only positively affects the flow of domestic tourists when it is used as a driver of organisation, coordination and integrated management of the tourism supply.

Natural disasters and university enrolment: evidence from L'Aquila earthquake. Applied Economics 49(14), 1440-1457, 2017. (with Giorgio Di Pietro) Working paper version

Abstract: Although there are several studies looking at the effect of natural disasters on economic growth, less attention has been dedicated to their impact on educational outcomes, especially in more developed countries. We use the synthetic control method to examine how the L’Aquila earthquake affected subsequent enrolment at the local university. This issue has wide economic implications as the University of L’Aquila made a large contribution to the local economy before the earthquake. Our results indicate that the earthquake had no statistically significant effect on first-year enrolment at the University of L’Aquila in the three academic years after the disaster. This natural disaster, however, caused a compositional change in the first-year student population, with a substantial increase in the number of students aged 21 or above. This is likely to have been driven by post-disaster measures adopted in order to mitigate the expected negative effects on enrolment triggered by the earthquake.

La valutazione delle politiche di incentivazione agli investimenti. EyesReg 5(6), November 2015.

If You Get Knocked Down, How Long Before You Get Up Again? International Journal of Sport Finance 9(4), 284-304, 2014.

Abstract: This paper addresses an important question on the consequences of relegation from some of Europe’s top football leagues: What is the team’s performance following relegation, compared to the situation without relegation? We compare the performance of relegated and non-relegated teams that battled until the last match to escape relegation in four large European leagues. We find that, on average, up to six seasons are necessary to completely reabsorb the negative relegation shock in sports outcomes. Additionally, we exploit current information on future TV rights revenues to forecast the extra cost of an ‘unlucky’ relegation in the 2013-14 season. The results show that, on average, relegation will cause an extra cost of about €135m to ‘combative’ teams relegated from the English Premier League. Smaller extra losses apply to the Italian Serie A (€60m), the Spanish La Liga (€38m), and the French Ligue 1 (€32m).

Conviene incentivare le piccole imprese piuttosto che le grandi? Un’analisi basata sul multiple regression discontinuity design. Rassegna Italiana di Valutazione a. XVIII - n. 58, 2014, 197-219. (with Guido Pellegrini)

Working Papers

Measuring Territorial Political Representation (with G. Pinto, A. Scerbo)

Abstract: Territorial representation is a cornerstone of democratic systems, ensuring that each geographic area has a voice in decision-making bodies. Although this principle is explicitly enshrined in federal governments, unitary states also incorporate territorial elements via mechanisms such as single-member district electoral systems and regional lists. However, certain democratic structures-most notably many large metropolitan governments-lack formal arrangements for territorial representation. This paper investigates whether, and to what extent, territorial disparities in political representation emerge in these contexts. To address this question, we propose several novel indices that measure the degree to which each territorial unit is represented relative to its population share. Using fine-grained precinct-level, geocoded electoral data, we calculate these indices for a sample of Italy's metropolitan areas and examine spatial and temporal variations in territorial representation. We find that areas with consistently below-average representation exhibit the lowest political engagement, the fewest university graduates, and the lowest employment rates. Our results contribute to the broader debate on the determinants and implications of territorial representation for democratic fairness and governance efficiency.

Signature Spending Politics (with C. Giannantoni, M. Letta, G. Pinto)

Abstract: We study the behavior of politicians when they have vast discretion over the allocation of public resources. We examine a unique case study from Italy, where the exogenous pandemic shock prompted Lombardy's regional government to approve a large extraordinary spending package. Most of these funds were swiftly allocated to numerous small projects through ad hoc motions proposed and signed by individual members of the regional parliament and subsequently approved through bipartisan resolutions. We leverage Large Language Models to text-mine thousands of parliamentary documents and assemble a fine grained politician-fund dataset containing detailed information on how and where these resources were distributed and which politician signed them. We document widespread pork-barrel strategies bargained with informal arrangements and driven by personal electoral rent-seeking, benefiting both majority and opposition politicians. By scraping social media data, we pinpoint the key mechanism: politicians credit claiming to boost visibility and advance their political careers.

'Dead Man Working': A Place-based Approach to Workplace Fatalities (with C. Giannantoni, M. Letta, G. Pinto)

Abstract: This paper proposes a place-based approach for addressing workplace fatalities. We employ machine learning (ML) techniques and a comprehensive panel dataset from Italy to detect systematic patterns that pinpoint areas most at risk of onthe-job fatalities. The empirical analysis demonstrates that ML algorithms can accurately forecast, ex-ante, the number of workplace deaths out of sample. We then create a granular risk map based on the ML forecasts and compare it with the actual allocation of on-site work inspections and public subsidies to improve occupational safety, finding minimal overlap. This mismatch suggests that public effort is currently not more prevalent where it is most needed, at least with respect to onthe-job deaths. Lastly, we assess the impact of on-site inspections on the number of workplace deaths via double/debiased machine learning and show that inspections appear to be effective only in areas flagged as high risk by the ML forecasts. Overall, these findings suggest that by replacing current allocation rules with machine predictions, it would be possible to significantly improve the cost-effectiveness of public interventions and boost the deployment of deterrent and preventive measures aimed at enhancing occupational safety and health.

The Clustered Dose-Response-Function Estimator for Continuous Treatment and Heterogenous Effects (with R. Di Stefano, R. Mattera)

Abstract: Many treatments are non-randomly assigned, continuous in nature, and exhibit heterogeneous effects even at identical treatment intensities. Taken together, these characteristics pose significant challenges for identifying causal effects, as no existing estimator can provide an unbiased estimate of the average causal dose-response function. To address this gap, we introduce the Clustered Dose-Response Function (Cl-DRF), a novel estimator designed to discern the continuous causal relationships between treatment intensity and the dependent variable across different subgroups. This approach leverages both theoretical and data-driven sources of heterogeneity and operates under relaxed versions of the conditional independence and positivity assumptions, which are required to be met only within each identified subgroup. To demonstrate the capabilities of the Cl-DRF estimator, we present both simulation evidence and an empirical application examining the impact of European Cohesion funds on economic growth.

Femicides, Anti-Violence Centers, and Public Salience: A Territorial Approach (with C. Giannantoni, M. Letta, G. Pinto).

Abstract: This paper conducts a comprehensive study on the geographic and socioeconomic patterns associated with femicides and examines the role of local public policies and saliency in combating gender-based violence. First, we compile a novel, granular dataset of femicide cases for Italy covering the period 2006-2022 and analyze it with machine learning techniques. This empirical analysis identifies areas at highest risk for women and pinpoints the main territorial predictors of the phenomenon. Second, we collect data on all local anti-violence centers (AVCs) and show that our femicide risk map does not fully align with the local deployment of existing public support, suggesting that machine predictions could be employed to refine targeting criteria. Third, using detailed information on the timing of AVC opening in each province and a staggered non-parametric matching difference-indifferences approach, we show that the establishment of AVCs did not significantly decrease the occurrence of femicides, while it lowered the number of reported sexual violence cases. Lastly, we present evidence that a widespread and exogenous increase in public and media salience of the phenomenon triggers large-scale short-term effects on helpseeking behavior that extend far beyond the local level. These findings suggest ample room for improving the targeting and effectiveness of policy interventions aimed at combating violence against women.

Causal inference and policy evaluation without a control group - October 2024 version (with Letta, M., Menchetti, F.) Replication Notebook. Media coverage: Interview by FLORENCE CENTER FOR DATA SCIENCE; LaVoce

Abstract: Without a control group, the most widespread methodologies for estimating causal effects cannot be applied. To fill this gap, we propose the Machine Learning Control Method, a new approach for causal panel analysis that estimates causal parameters without relying on untreated units. We formalize identification within the potential outcomes framework and then provide estimation based on machine learning algorithms. To illustrate the practical relevance of our method, we present simulation evidence, a replication study, and an empirical application on the impact of the COVID-19 crisis on educational inequality. We implement the proposed approach in the companion R package MachineControl.

Pay Incentives in Politics: Evaluating a Large-scale Salary Increase for Local Politicians (with S. Nocito, G. Pinto). Media coverage: Youtrend

Abstract: We evaluate the impact of a recent reform that sharply increased the salaries of Italian local politicians on electoral competition and the valence attributes of the candidates elected. Exploiting misaligned election dates across Italian cities, we propose a novel methodology, the shifted difference-in differences design (Sh-DiD), to estimate the reform’s impact on municipalities up to 30,000 inhabitants, representative of almost the entire universe of Italy’s local administrative units. We find a boost in the entry of new political candidates after the first post-reform electoral round, with no significant enhancement in the overall quality of the political class. These outcomes possibly stem from the varying distribution of compliers whose candidacy decision is influenced by the reform—across diverse political and economic contexts. Thus, we find that in less affluent areas or those with fewer entry barriers, the pay rise drew a larger number of mayoral candidates, encouraging individuals from outside the political sphere to enter the competition. In the poorest contexts, we also observe a shift in the profile of councilors and members of the mayor’s executive committee, where the pay rise attracted individuals with lower educational levels but with experience in white-collar positions.

Income tax rate increases and heterogeneous taxpayers’ reactions: a spatial regression discontinuity design. (with Emma Galli) DISSE Working Papers Series No. 17/2020, 2020.

Abstract: This paper exploits a sudden increase in the regional surcharge on the income tax rate in Lazio, one of the most populated regions of Italy, to compare taxpayers’ reported income in Lazio’s municipalities with those in the municipalities located in six neighboring regions. To this end, we have built a new yearly dataset (2012-2018) at municipal level containing the reported income of different categories of taxpayers and employ a spatial regression discontinuity design to estimate the response to an increase of the marginal tax rate in terms of reported taxable income by different types of taxpayers. We find a sizable and persistent decrease in reported income only for the self-employed and entrepreneurs, while the employees respond by slowly reducing their declared incomes. As expected, the retirees do not exhibit any response.

Work in Progress

Going Local: Public Spending, Bureaucratic Efficiency and Decentralization (with C. Giannantoni)

Labor mobility in times of turmoil: A comparative analysis of the effects of different economic crises on displaced workers (with V. Celli, G. Pellegrini)