On February 17, 1992, the arrest of Mario Chiesa, a socialist administrator at the Pio Albergo Trivulzio, a rest home in Milan, marked the first rift in the Italian political system. What seemed like an inconsequential event unfolded into a significant turning point, triggering a sequence of arrests involving politicians from various government parties. This unfolding scenario gradually unveiled a pervasive network of corruption that extended beyond mere isolated cases of bribery. Indeed, it exposed a sophisticated web of corrupt practices involving collusion between private company cartels, political officials, and intermediaries, and even connections to organized crime. The rapid involvement of prominent politicians in the sweeping investigation known as “Mani Pulite” (“Clean Hands”)^Footnote1 made it the largest probe in Italian political history. The enduring image of coins being hurled at Bettino Craxi, the former Prime Minister, outside the Hotel Raphael in Rome during the spring of 1993 remains ingrained in the collective imagination.

Beyond being mere anecdotes, these events underscore how corrupt politicians in Italy gained and sustained support by establishing a system of patronage to oversee public affairs. In the four decades leading up to these events (1950–1990), the public system played a pivotal role, with politicians holding a fundamental position within it. A recent study conducted by Barone et al. (2022) investigated the political connection premium across Italian cities during this 40-year period, spanning from the aftermath of World War II to the fall of the Berlin Wall. Their findings revealed that politically connected cities—defined as the birthplaces of influential politicians and neighboring municipalities—experienced a population premium of approximately 8% over four decades. This premium was to a significant extent attributed to investments in infrastructure and the strategic location of state-owned enterprises, both contributing to industrialization, heightened employment rates, and increased wages.

Several other studies have similarly highlighted the pivotal role of politicians in bolstering local economic growth, exploring regional favoritism, and examining the broader impact of politicians on hometown development within both democratic and autocratic frameworks. Significant contributions include works by Hodler and Raschky (2014), Kahn et al. (2021), Do et al. (2017), Gonschorek et al. (2018), and Dreher et al. (2019), among others. The political connection premium has also been scrutinized at the European level (Gehring and Schneider 2018), as well as within various individual European nations (Fiva and Halse (2016) in Norway, Baskaran and da Fonseca (2021) in Germany, Lévêque (2020) in France, Carozzi and Repetto (2016) and Golden and Picci (2008) in Italy).^Footnote2

In this paper, I undertake an analysis to ascertain the persistence of the political connection premium for Italian cities in the post-Mani Pulite era, encompassing the two decades from 1991 to 2011. To the best of my knowledge, no prior investigations have examined whether the premium still exists after 1991. The matter is particularly intriguing given that during the so-called “First Republic” (1950–1990), politicians retained positions of power for extensive periods, coinciding with a prevalent system of widespread corruption.^Footnote3 During this era, the state’s significant involvement in the economy was evident. Subsequently, at the onset of the 1990s, substantial transformations took place. The Mani Pulite scandal, coupled with the introduction of a new electoral law, profoundly reshaped the political landscape. This period marked the establishment of a rotational mechanism between the majority and opposition, leading to the emergence of a more pronounced bipolar system compared to the preceding four decades. This new system cyclically involved two coalitions: the center-right and center-left (Pasquino and Valbruzzi 2015). The dissolution of long-standing dominant parties within the political sphere over the previous 40 years led to a noteworthy extension in the average duration of governments. Nonetheless, a degree of continuity with the preceding era persisted, as several ministers, undersecretaries, and politicians maintained pivotal roles across successive governments.

Regarding the public sector’s significance within the economy, contrasting scenarios emerged in the early 1990s. While initiatives like the divestiture of state holdings, the privatization of public entities, and the liberalization of economic sectors and public services were set in motion, the public sector continued to maintain a substantial role in the economy. Specifically, the portion of public expenditures in gross domestic product (GDP), initially standing at 52.6% in 1991, diminished to 46.5% by 2000. However, it subsequently underwent an upswing in the following decade, stabilizing at around 50% by 2011 (Eurostat data). Despite experiencing fluctuations, this percentage consistently remained at a high level. Additionally, a process of fiscal decentralization, notably through Bassanini’s reform in 1997 and the constitutional reform in 2001, led to a new functional composition of public expenditure (Ambrosanio et al. 2010; Grisorio and Prota 2015).^Footnote4

In this context, I outline a cohort of influential politicians, encompassing incumbent prime ministers and leaders of political parties who held governmental roles or provided external support^Footnote5 during the 1991–2011 period. Subsequently, I investigate whether the municipalities of these politicians’ hometowns and other nearby municipalities (within a 10-km radius^Footnote6 of the birth city of the aforementioned politicians)—treated municipalities—exhibited more robust economic growth during the two decades of the so-called “Second Republic” (1991–2011) compared to similar untreated municipalities. To accomplish this, I conduct a fixed effects regression analysis, correlating the population growth rate at the municipal level between 1991 and 2011 with a treatment indicator and a full set of municipal-level characteristics. Conditional on these observable predetermined characteristics, the identification assumption asserts that the treatment status is effectively randomly assigned. Hence, this research framework enables an evaluation of the influence of political affiliations on the population growth rate at the municipal level.^Footnote7 It is noteworthy that this analysis specifically targets small and medium-sized cities, deliberately excluding large cities and their neighboring municipalities to mitigate potential confounding variables that could distort the analysis by influencing population growth in larger urban areas. Moreover, in an effort to establish a control group that closely parallels the treatment group, I omit from the control group municipalities with the lowest likelihood of being treated, as suggested by the control variables.^Footnote8

The estimates indicate the existence of a political connection premium among treated municipalities, showing an approximately 2.5% higher population growth rate with respect to control municipalities over a span of 20 years (equivalent to 12% of the standard deviation of the dependent variable). A visual inspection of municipal population growth within both the treatment and control groups in the two decades preceding the treatment period validates the assumption of parallel trends.^Footnote9 The results hold up under several robustness checks,^Footnote10 including a replicated analysis excluding mid-large sized cities^Footnote11 and politicians holding office prior to 1994 (the year of the first post-Mani Pulite general elections).^Footnote12 Additionally, the findings remain robust across sensitivity analyses such as trimming the dependent variable, altering the control sample,^Footnote13 including large cities in the control group, changing the threshold (in kilometers) defining a city in the treatment group, and dropping from the analysis the birthplaces of politicians from the First Republic (and neighboring municipalities). Furthermore, the results remain robust and show a large magnitude when focusing solely on a subset of politicians—former prime ministers, party leaders in government or providing external support, but also members of parliament with a minimum of four mandates. The persistence of the political connection premium remains evident even when employing different eligibility criteria for politicians, such as requiring politicians to have held either a ministerial or an undersecretary to the prime minister position at least once within the two decades.^Footnote14 The findings also indicate that politicians who have been in office for a longer period exert a more substantial influence on the municipalities connected to them compared to those more recently appointed.^Footnote15 Diverse clustering methodologies and propensity score matching estimates yield consistent results without significant alterations. Moreover, the analysis holds true when considering the hometowns of politicians and their surrounding areas across subsequent governments, shifting the temporal perspective to different timeframes, including the periods of 1991–2014, 1991- 2016, and 1991–2019.^Footnote16

This study contributes to the existing literature along several dimensions. Firstly, it unveils that in Italy, the phenomenon of a “political connection premium” was not confined solely to the First Republic but, rather, persisted through the Second and Third Republics (1991–2019). Secondly, through an exploration of diverse social capital and institutional quality indicators, I discern that these factors play a pivotal role as mediators in mitigating the effects of political connections.^Footnote17 Thirdly, an analysis of municipal-level firm data (ASC—Statistical Atlas of Municipalities 2012) illustrates that politically connected municipalities exhibit a higher prevalence of medium, medium-large, and large firms, coupled with fewer employees per small firms. This implies that treated (politically connected) municipalities might have received stronger incentives to attract medium-to-large firms and, conceivably, more generalized incentives for small firms compared to control municipalities. These findings are in line with the analysis of economic structures, showcasing the exceptional performance of treated municipalities concerning broadband connectivity, employment rates (especially skilled employment), and the presence of university-educated individuals.^Footnote18 Fourthly, leveraging granular income data from the Ministry of Economic and Financial Affairs (MEF, 2000–2019), I employ information on disposable income as an additional proxy for local development, complementing and confirming the analysis of population growth rates.^Footnote19 Fifthly, contrary to the trends observed in the preceding forty years, the analysis indicates that spillover effects appear robust and not geographically concentrated. This may be partially attributed to enhancements in local public transport and the ratio of cars per inhabitant.^Footnote20

Finally, based on data from the Ministry of the Interior (MINT) on the 2013 and 2018 general elections, I discern a two-way relationship: descriptive evidence delineates a positive influence of political connections on voter turnout and a negative effect concerning votes for populist parties (in 2013). It is crucial to interpret this phenomenon cautiously as a potential consequence of political connections, resembling a reciprocal (“do ut des”) relationship.^Footnote21

This paper is structured as follows. Section 2 provides a concise literature review. Section 3 delves into the social, political, and economic landscape during the examined period (1991–2011). Section 4 presents the empirical framework, explores the institutional background, the construction of the empirical strategy, and the definitions of the treatment and control groups, discusses the data, provides descriptive statistics, and outlines the identification strategy. Section 5 details the empirical results, along with placebo and robustness checks, and analyzes heterogeneous effects. Section 6 investigates structural changes and delves into the topics of employment and firms, while Sect. 7 scrutinizes the impact of political connections on political outcomes. Section 8 discusses the conclusions drawn from the study. Appendix 1 includes additional robustness checks, supplementary figures and tables, and a list of governments and parties from the 1991–2019 period. Finally, Appendix 2 provides an in-depth analysis of spillover effects.

In recent years, an extensive literature has addressed the issue of political connection premiums, both at the global level (Hodler and Raschky (2014) and De Luca et al. (2018)) as well as with a focus on autocracies and democracies in various parts of the world, applying several methods and different levels of granularity (China: Kahn et al. (2021), Kung and Zhou (2021), and Jiang and Zhang (2020); Vietnam: Do et al. (2017); Turkey: Zhang (2021); Africa: Dreher et al. (2019), Burgess et al. (2015), and Franck and Rainer (2012); India: Khalil et al. (2021); Latin America: Mattos et al. (2021) and Firpo et al. (2015)). With respect to Europe as a whole, Gehring and Schneider (2018) analyzed data on EU agriculture commissioners, showing that the nationality of EU commissioners influences budget allocation decisions, clearly in favor of the commissioner’s home nation.^Footnote22 On average, they estimate that the home nation of the agriculture commissioner gets an extra 1% of resources in relation to the overall EU budget, which corresponds to just under €1 billion annually (€850 million).

However, when it comes to Europe, most papers take a national perspective. Concerning northern Europe, Fiva and Halse (2016) employed data from the norwegian regional governments to investigate the norwegian case, characterized by a closed-list proportional representation system where politicians lack a clear electoral incentive to prioritize their hometown. Nevertheless, they found a hometown bias: municipalities with a representative in the regional council from the same party as the regional governor receive more funding for local investment. As the authors point out, this finding suggests that electoral incentives are not the sole factor driving so-called “pork barrel” politics; social ties and group identities among politicians also seem to play a role.

Baskaran and da Fonseca (2021) conducted a study on the German case. They matched the hometown data of ministers from various German states with government employment data spanning the period from 1994 to 2013. They found that cities in which at least one minister resides who has been in office for more than one term experience a higher annual growth rate in government employment, including police officers, teachers, and administrative staff. Consequently, their results suggest that appointed cabinet members engage in hometown favoritism. In their opinion, politicians do not exhibit this behavior to maximize their chances of reelection, as the proportional system does not incentivize such conduct. And this is also true for any other political interests such as municipal elections, given the way political careers are made in Germany. Therefore, the authors propose that the driving force behind this behavior is the desire to enhance the provision of public goods in their cities, assist friends and family, and reward local party members, as well as an inability to resist job requests from fellow residents.

Lévêque (2020) studied the influence of political connections on public policies implemented at the local level in France. He examined a sample of nearly 200,000 local politicians in French municipalities with over 3500 inhabitants between 2008 and 2014. His findings indicate that households of candidates who supported the elected mayor got 35% more building permits than households supporting a defeated mayoral candidate during the mayor’s 6-year term. This effect was particularly pronounced in cities characterized by low political competition and was diminished when there were close elections.

Among others, Carozzi and Repetto (2016) have looked at the situation in Italy. Using a panel of Italian cities over the period of 1994 to 2006, they found that the municipal governments of legislators’ birthplaces receive larger transfers per capita than others. Additionally, exploiting variation in birthplaces induced by parliamentary turnover, they found that the effect is driven by legislators who were born in a town outside of the district in which they were elected. This result thus elucidates how such copious transfers cannot be explained by politicians’ re-election incentives, the traditional explanation used in the literature for so-called pork-barrel transfers, but is instead driven by other personal reasons. Indeed, the amount of birthplace bias of politicians evolves over time, being highest in the years following elections and disappearing at the end of their mandate, when politicians are focused on their re-election. The authors identified at least a partial explanation: politicians enhance their prospects of being elected at the local level once their parliamentary tenure concludes by increasing funding toward their birthplace. Furthermore, they demonstrate that this mechanism is more pronounced when the mayor in the receiving municipality and the parliamentarian belong to the same political party.

Golden and Picci (2008) analyzed the political determinants of the Italian government’s distribution of infrastructure funds to the NUTS-3 regions between 1953 and 1994. They found that districts that elect more powerful individuals—regardless of whether they are on the lists of the governing parties—secure more investment in infrastructure. This means that in the open-list electoral environment that prevailed in post-war Italy, the most powerful deputies were able to secure resources at the expense of the governing parties themselves.

The closest work to the current study is that of Barone et al. (2022). They estimated the political connection premium for Italian municipalities during the latter half of the twentieth century, primarily focusing on the political landscape of the so-called First Republic period, spanning from World War II to the fall of the Berlin Wall. To do so, they compared the population growth rates of connected cities with those of similar but unconnected cities. Their findings reveal that politically connected municipalities experienced a population premium of 8% over a span of 40 years. The authors document how politicians’ hometowns benefited from increased infrastructure investments and the establishment of facilities by state-owned enterprises. Consequently, this has led to heightened industrialization, employment, and wages in these areas. Thus, Barone et al. (2022) underscore that the growth premium associated with political connections is not solely confined to developing economies, as suggested by prior research. Moreover, their analysis encompassed a substantial 40-year period and, as a result, it is more likely that they were able to capture the steady-state spatial equilibrium.

The twenty-year period under consideration here spans from 1991 to 2011, marking a significant shift from Italy’s First Republic, which lasted from the conclusion of World War II until 1994. Following World War II, Italy transitioned to a republic through the 1946 referendum and enacted a new republican constitution.

Over the subsequent 40 years, the political system exhibited stability, partly owing to the influence of the Cold War and the use of a proportional electoral system. During this time, there were numerous changes in government, with a total of 45 shifts between 1948 and 1992. Nonetheless, it is important to note that many of these transitions primarily involved the reshuffling of ministerial positions among various political factions, rather than representing substantial alterations. Five major parties dominated the political scene, holding around 60% of the national vote: DC, PSI, PSDI, PLI, and PRI.^Footnote23 The PCI (Italian Communist Party) served as the leading opposition party. This era has been labeled a “partitocracy” or a time of “party politics” (Crainz 2003) due to the concentration of power within the parties themselves. Throughout the 1980s, there was an aspiration for a transformation mirroring the structural alterations implemented by Charles De Gaulle in France, foreseeing a transition towards what was conceived as a “Second Republic”.^Footnote24 This envisioned shift did not come to fruition; instead, the alterations predominantly occurred within the established contours of the existing political system. The only change that could affect the institutional setup was the changing of the electoral law in a majoritarian sense.^Footnote25 In any case, the early 1990s brought profound changes, both politically and socioeconomically. Issues such as illegal immigration, intolerance, economic stagnation, loss of competitiveness, high inflation, and a growing public deficit shaped the socioeconomic landscape.

In addition, the fall of the Berlin Wall coincided with the transformation of the PCI (Italian Communist Party) into the Democratic Party of the Left, realigning the Italian Left. Regionalist movements in northern Italy, such as the Lombard League, emerged in opposition to centralization and the established party system, leading to further fragmentation. In the south, criminal organizations had significant influence, particularly the Mafia, which was at its peak in the early 1990s and ordered the assassination of prominent magistrates. However, the most significant disruption came with the Mani Pulite (Clean Hands) scandal that implicated numerous politicians in corruption. It exposed a system of illegal political financing involving private entrepreneurs, primarily affecting the DC (Christian Democratic Party) and the PSI (Italian Socialist Party), the two dominant parties of the period (Sabatucci and Vidotto 2008). The political landscape underwent a profound transformation as the major political parties saw a near-complete loss of their votes^Footnote26 and businessman Silvio Berlusconi entered the political arena by founding Forza Italia. In fact, the 1994 elections marked a watershed between the First and Second Republics. In the following years, the political landscape evolved into a bipolar system with two main coalitions, “L’Ulivo” (later “L’Unione”) representing the center-left and “Il Polo delle Libertà” (later “Casa delle Libertà”) representing the center-right. This balance largely persisted until 2011, except for a brief shift in 2008 when the “Partito Democratico” (center-left) and “Il Popolo della Libertà” (center-right) emerged.

The real game changer came in 2013 with the entrance of the Five Star Movement onto the political scene, which effectively marked the crisis of bipartisanship (Pasquino and Valbruzzi 2015). However, over these two decades, the average duration of governments increased and dialogue between majority and minority parties heightened. Throughout this period, the state continued to play a significant role in the economy, despite privatization and liberalization efforts that began in the 1990s, and corruption remained a persistent issue.

Furthermore, it is noteworthy that during those years Italy was undergoing a process of institutional and fiscal decentralization. Such a process had already affected many countries, both advanced and developing (Rodríguez-Pose and Gill 2003), and in European countries it was driven by the principle of subsidiarity incorporated into the Treaty of Maastricht. A significant step in this direction was the Bassanini reform in 1997, which promoted the decentralization of administrative functions, followed on the financial side by the establishment of the IRAP (Imposta Regionale sulle Attività Produttive, a tax on value-added raised at the firm level).

Subsequently, the 2001 constitutional reform, which granted greater exclusive competencies to the regions, led to a new functional composition of public expenditure (Ambrosanio et al. 2010; Grisorio and Prota 2015).^Footnote27