OVERS

Countries around the world have launched public disclosure programs to stimulate citizen participation in environmental governance, yet little is known about whether such participation is effective and if it is what makes it so. We layer a national-scale field experiment that randomly files pollution appeals to either regulators or the violating firms through public or private channels. We find that, publicly appealing to the regulator of a firm’s violation on popular social media increases both regulatory oversight and firm compliance, which reduce subsequent violations by 40% and air and water pollution emissions by 12% and 5%, respectively. In contrast, appealing to the regulator through private channels only causes a small and statistically insignificant improvement in environmental outcomes. Additionally, we randomly vary the proportion of firms subject to appeals at the prefecture-level and find that there is a positive general equilibrium impact as the control firms in high-intensity prefectures reduce violations more than control firms in low-intensity prefectures. Analysis of ambient pollution data and additional back-of-the-envelope calculations both suggest that encouraging public participation in environmental governance would lead to significant improvements in China’s aggregate environmental quality.

We examine whether news media compromise their reporting to maintain market access in authoritarian countries by exploiting a large-scale media crackdown in China in May 2019, in which multiple influential UK- and US-based news sites were blocked. We find that outlets adopted a more negative tone toward China after losing access, compared to those with no access change. The negative effect is present in news on politically sensitive topics only — not economic topics — and is stronger for media outlets with less influence in China. Further evidence suggests that these findings may be interpreted as the media censoring themselves less after being blocked.

This paper studies the effects of immigration on crime and crime perceptions in Chile, where the foreign-born population more than doubled in the last decade. By using individual-level victimization data, we document null effects of immigration on crime but positive and significant effects on crime-related concerns, which in turn triggered preventive behavioral responses, such as investing in home-security. Our results are robust across a two-way fixed effects model and an IV strategy based on a shift-share instrument that exploits immigration inflows towards destination countries other than Chile. On mechanisms, we examine data on crime-related news on TV and in newspapers, and find a disproportionate coverage of immigrant-perpetrated homicides as well as a larger effect of immigration on crime perceptions in municipalities with a stronger media presence. These effects might explain the widening gap between actual crime trends and public perceptions of crime.

Did mobile Internet contribute to the rise of populist, anti-elitist and non-traditional parties across Europe? We investigate these questions using geo-referenced data on 3G and 4G mobile phone signal coverage, together with geo-referenced national electoral results across over 80,000 municipalities in 25 European countries between 2006 and 2018. We complement our analysis with survey data from the European Social survey.

This paper provides causal evidence on how partisan news messaging from cable television influences the content published by newspapers in U.S. localities. We introduce a new parallel corpus of newspaper articles (24M articles in 600+ local newspapers) and transcribed television news shows (40K cable news episodes from Fox News Channel CNN, and MSNBC) for the years 2005-2008. We measure media influence using a supervised learning model that predicts, for a given piece of text, the probability that it comes from a Fox News transcript, rather than from CNN or MSNBC. After validating the measure, we apply it to the local newspaper article texts. Exogenous variation in news viewership across localities comes from relative channel numbering, which we use as instruments. We find that an exogenous increase in local viewership of a cable news network shifts the textual content of local newspapers toward that network’s content. Televised media slant works not just through persuading viewers, but through influencing other media outlets.

10:00-10:45 Jörg Spenkuch (Northwestern) Ideology and Performance in Public Organizations (with Edoardo Teso and Guo Xu)

10:45-11:30 Kathryn Baragwanath (UCSD) The Political Economy of Deforestation in the Brazilian Amazon

11:30-11:45 COFFEE BREAK

11:45-12:30 Gabriele Gratton (UNSW) Liberty, Security and Accountability: The Rise and Fall of Illiberal Democracies (with Barton E. Lee)

12:30-2:45 LUNCH BREAK

2:45-3:30 Shawn Treier (ANU) Founding Factions: How Majorities Shifted and Aligned to Shape the U.S. Constitution (with Jeremy C. Pope)

3:30-4:15 Umair Khalil (Monash) Gender, Discouragement, and the Stigma of Failure (with Marco Faravelli and Sundar Ponnusamy)

4:15-4:30 COFFEE BREAK

4:30-5:15 Bård Harstad (Oslo) The Conservation Multiplier

5:15-6:00 Alessandra Casella (Columbia) Minority Turnout and Representation under Cumulative Voting. An Experiment (with Michelle Jiang and Jeffrey Guo)

Telephone operation, one of the most common jobs for young American women in the early 1900s, provided hundreds of thousands of female workers a pathway into the labor force. Between 1920 and 1940, AT&T adopted mechanical switching technology in more than half of the U.S. telephone network, replacing manual operation. We show that although automation eliminated most of these jobs, it did not affect future cohorts' overall employment: the decline in demand for operators was counteracted by growth in both middle-skill jobs like secretarial work and lower-skill service jobs, which absorbed future generations. Using a new genealogy-based census linking method, we show that incumbent telephone operators were most impacted by automation, and a decade later were more likely to be in lower-paying occupations or have left the labor force entirely.

Can politics motivate propaganda in media? This paper examines the case of the unexpected and short-lived electoral success of the pro-redistribution Populist Party in the 1892 presidential elections. The Populists sought support among poor farmers, regardless of race. This biracial alliance threatened the Democratic establishment in the South, providing it with an incentive to fan racial fears to split the newly formed coalition. Newspapers affiliated with the Democrats spread propaganda of attacks by Blacks on the White community, often involving allegations of sexual assault. Using novel newspaper data, we identify these hate stories and show that they become more prevalent in the years following the 1892 presidential election in counties where the Populists were active. The effect is large and found in newspapers affiliated with the Democrats only. The evidence suggests that the propaganda ``worked'': where newspapers spread more propaganda, the Democrats see stronger gains in presidential elections in the following decades, long after the Populists left the political arena.

We develop a model of electoral accountability with mainstream and alternative media. In addition to regular high- and low-competence types, the incumbent may be an aspiring autocrat who controls the mainstream media and will subvert democracy if retained in office. A truthful alternative media can help voters identify and remove these subversive types while re-electing competent leaders. A malicious alternative media, in contrast, spreads false accusations about the incumbent and demotivates policy effort. If the alternative media is very likely be malicious and hence is unreliable, voters ignore it and use only the mainstream media to hold regular incumbents accountable, leaving aspiring autocrats to win re-election via propaganda that portrays them as effective policymakers. When the alternative media's reliability is intermediate, voters heed its warnings about subversive incumbents, but the prospect of being falsely accused demotivates effort by regular incumbents and electoral accountability breaks down.

We examine how fiscal legibility, the ability of a central government to observe local economic conditions for the purposes of taxation, shapes political centralization. When a ruler is unable to observe economic conditions, it can be preferable to grant autonomy to local intermediaries in charge of tax collection to encourage better performance. As a ruler's ability to observe local conditions improves, so does his ability to monitor and sanction of underperforming intermediaries. This enables the ruler to tighten control over tax collection, retain more revenue, and establish a more direct state presence. This shift also encourages the ruler to invest in further enhancing fiscal legibility over the longer term. We present a dynamic principal-agent model to illustrate this argument and provide empirical support for the theory using subnational panel data on local political institutions in colonial Mexico. We focus on the effects of a technological innovation that drastically improved the Spanish Crown's ability to observe local economic production: the introduction of the patio process to refine silver. We show that the transition to direct rule differentially increased in mining districts following this technological innovation and that these areas saw greater investments in improving state informational capacity over the long term.

This paper studies a dynamic model of property ownership and appropriation in modern autocracies. An autocrat represents the interests of an elite "in-group." It chooses whether and how much to appropriate from public and from private assets of an "out-group" at each date. To maintain the appearance of the rule of law, the autocrat proposes an assignment of ownership that takes effect only if it is accepted by the affected citizens. However, the autocrat's commitment to an accepted assignment is only temporary, and because its enforcement of property rights is tied to the duration of this commitment, the autocrat can systematically appropriate property and wealth from the out-group and from public assets. Under some initial conditions, the autocrat initially implements popular land reform only to take it back down the road. More generally, the wealth shares of both public property and private property of the out-group decline monotonically after an initial adjustment period. The model rationalizes the connection between increasing wealth inequality and privatization occurring in many autocracies. The dynamics of wealth and appropriation are consistent with Russian and Chinese wealth data, and simulations of these countries' wealth distributions to mid 21st century display widening gaps in wealth between elites and the rest of the populace. Finally, we show that the ruling group under anocracy, an autocratic system that admits opposition and civil society groups, will generally be better off than under a traditional autocracy. The dilemma is that the anocratic system might enable the growth of an opposition party that eventually displaces the ruling group.

Monitoring is a common tool used to mitigate agency problems. Monitors themselves, however, may be biased or corrupt, in particular if they feel obliged to please the party that appoints them. In this paper, I evaluate whether shifting control over auditor assignments improves monitoring effectiveness and impacts outcomes of the audited party. In 2011, Italy switched from allowing mayors to appoint municipal auditors to a system of random assignment, to strengthen oversight and ensure the financial soundness of municipal budgets. My identification exploits the reform's staggered introduction across municipalities in a generalized difference-in-differences setting. I obtain three main findings. First, treated municipalities increase their net surpluses by 9% and debt repayments by 8%, in accordance with national government objectives. Second, the improvement is achieved through revenue-based consolidation, rather than by cutting expenditures. Third, treatment effects are significantly larger for municipalities that were more at risk of collusion before the reform, and for those that are matched to a more distant or less connected auditor. Taken together, these findings provide novel quantitative evidence on the importance of independence in auditing, and highlight the improvement in outcomes that may result from changes in the design of monitoring institutions.

Oil royalties provide a substantial but volatile inflow of non tax-payer money to municipal coffers. While a large literature examines the impact of oil on democratic emergence and stability, I examine how oil impacts corruption and the types of candidates elected under democracy. To predict the effects of oil royalties, I develop a formal model with moral hazard, adverse selection and endogenous entry. I show that natural resource windfalls generate the strategic entry of corrupt candidates and prevent voters from distinguishing politicians' integrity, creating cycles of corruption and reelection. I test this theory in Brazil, where offshore royalties are determined and allocated exogenously based on a geographic rule and the international price of oil. Consistent with the model, I find that a one standard deviation increase in oil royalties produces a 29% increase in corruption. The effects of windfalls on corruption are larger after elections during booms and lower during busts. Furthermore, oil royalties lead to a reelection cycle: when the price of oil is expected to be higher, incumbents are reelected more often than when the price of oil is expected to fall, independent of economic and individual level variables. I show that strategic entry of corrupt candidates during booms is likely the cause of these corruption and reelection cycles, as predicted by the theory. Taken together, these results point to a strong effect of oil royalties on local level corruption and electoral outcomes.

COVID-19 caused a significant health and economic crisis, a condition identified as conducive to stigmatization and hateful behavior against minority groups. It is however unclear whether the threat of infection triggers violence in addition to stigmatization, and whether a violent reaction can happen at the onset of an unexpected economic shock before social hierarchies can be disrupted. Using a novel database of hate crimes across Italy, we show that (i) hate crimes against Asians increased substantially at the pandemic onset, and that (ii) the increase was concentrated in cities with higher expected unemployment, but not higher mortality. We then examine individual, local and national mobilization as mechanisms. We find that (iii) local far-right institutions motivate hate crimes, while we find no support for the role of individual prejudice and national discourse. Our study identifies new conditions triggering hateful behavior, advancing our understanding of factors hindering migrant integration.

New technologies make it easier for charismatic individuals to influence others. This paper studies the political impact of the first populist radio personality in American history. Father Charles Coughlin blended populist demagoguery, anti-Semitism, and fascist sympathies to create a hugely popular radio program that attracted tens of millions of listeners throughout the 1930s. Exploiting variation in the radio signal strength as a result of topographic factors, I find that a one standard deviation increase in exposure to Coughlin’s anti-Roosevelt broadcast reduced FDR’s vote share by about two percentage points in the 1936 presidential election. Effects were larger in counties with more Catholics. An alternative difference-in-differences strategy exploiting Coughlin’s switch in attitude towards FDR during 1932-1936 confirms the results. Moreover, I find that places more exposed to Coughlin’s broadcast in the late 1930s were more likely to form a local branch of the pro-Nazi German-American Bund and sold fewer war bonds during WWII.

We consider the effects of climate change on seasonally migrant populations that herd livestock – i.e., transhumant pastoralists– in Africa. Traditionally, transhumant pastoralists benefit from a cooperative relationship with sedentary agriculturalists whereby arable land is used for crop farming in the wet season and animal grazing in the dry season. Droughts can disrupt this arrangement by inducing pastoral groups to migrate to agricultural lands before the harvest, causing conflict to emerge. We examine this hypothesis by combining ethnographic information on the traditional locations of transhumant pastoralists and sedentary agriculturalists with high-resolution data on the location and timing of rainfall and violent conflict events in Africa from1989–2018. We show that droughts in the territory of transhumant pastoralists lead to conflict in neighboring agricultural areas. Additionally, (i) the conflict is concentrated in the wet season and not the dry season; and (ii) the mechanism operates through rainfall’s effect on plant biomass growth. We also find that this effect on conflict is greater in countries where pastoral groups have less political power. The magnitudes of our estimates indicate that nearly all of the reduced-form relationship between adverse rainfall shocks and conflict in Africa is explained by this mechanism.

We provide new evidence on how conflict adversely affects economic outcomes. Specifically, we ask whether and how the production network is a first-order determinant of the propagation of conflict to firms outside of conflict zone. Using microdata on Indian manufacturing plants and geo-coded information on Maoist insurgency, we first provide an estimate of the direct exposure to conflict. Firms located in conflict affected areas suffer a loss of 7-11% of their output. Estimating structurally a general equilibrium model of production networks, we then obtain an estimate for the overall macroeconomic impact of the Maoist insurgency by taking this propagation effects into account. We find that the Maoist insurgency resulted in a 0.4-0.7% decline in aggregate output of Indian’s manufacturing sector. Only the 20% of this loss is due to direct exposure to conflict, whereas the remaining 80% explained by the indirect exposure to conflict through the network production.

Why does interpersonal violence persist in the US at high rates? Some historians have argued that the relatively late development of effective state institutions in parts of the US forced people to handle their own defense, acting preemptively and taking justice into their own hands. We show that legacies of violence from these weak-state areas appear to spread and persist across the US as whites migrated from these regions to other states. The risk of homicide victimization white migrants faced in their birth states largely explains their victimization risk in their ultimate state and county of residence, even when accounting for alternative explanations, such as economic differences and social disorganization.

In the United States, black workers work in larger establishments than white workers. We document a similar pattern in Brazil: the nonwhite share of the formal sector workforce is increasing in establishment size. We show that this pattern is not driven by differences in job characteristics or skill requirements between establishments. We hypothesize that larger employers hire more nonwhite workers because they invest in formal screening methods and are less dependent on informal methods that tend to be more informative for candidates from advantaged groups. We build a model of firm hiring that features this mechanism and derive predictions. Consistent with the model, we find that: (1) at the establishment level, the nonwhite share of hires is increasing in hiring volume; (2) nonwhite-white within-establishment differences in separation rates are declining in hiring volume; (3) between-establishment segregation is declining in hiring volume.

The link between time of day and productivity on cognitive tasks is crucial to understanding workplace efficiency and welfare. In this paper, we aim to answer a general question and investigate how time-of-day affects cognitive tasks performance exploiting an ideal setting in which tasks are assigned at different time of day in a quasi-random fashion. Specifically, we studied the performance of university students sitting no more than one examination per day in the final two weeks of a semester. Using half million observations on academic records, we found that students perform better at lunchtime (1.30 pm) than in the morning (9 am) or late afternoon (4.30 pm). This inverse-U shape relationship between time of day and performance (i) is not driven by stress or fatigue, (ii) is consistent with the idea that cognitive functioning is an important determinant of productivity, and (iii) implies that efficiency gains of up to 0.14 standard deviations can be achieved by simply re-arranging examination times. A simple back of the envelope calculation applied to an external context that is likely to benefit from our results (i.e., elective surgeries), suggests that sorting the cognitive tasks performed by surgeons differently could lead to an increase in the number of patients saved.

Given the importance of social capital to economic outcomes, in this paper we ask what shapes social capital in the long run. In particular we study if the historical shock of the Indian Partition (1947) affects social cohesion and trust in Indian districts sixty years later. We measure the Partition shock as the proportion of “displaced” migrants in an Indian district in 1951. India was partitioned in 1947 into India and Pakistan (East and West Pakistan). At this time, many Hindus and Sikhs migrated from Pakistan to India while Muslims went in the other direction. Using data from the World Health Organization Survey on the Aged and Elderly conducted in six states of India, we find strong evidence that social cohesion and trust is lower in districts that received more Partition migrants. The effects remain robust in presence of spatial robustness checks as well as controlling for other contemporaneous factors, public goods provisions, Hindu-Muslim riots, literacy, urbanization and the gender ratio. Rather, these effects are mediated through later waves of migrations: districts with more Partition migrants seed the subsequent growth of migrants especially in 1971 – 1981, driven by wars on the Eastern and Western borders between India and Pakistan.

Political polarization has ruptured the fabric of U.S. society. The focus of this paper is to examine various layers of (non-)strategic decision-making that are plausibly affected by political polarization through the lens of one's feelings of hate and love for Donald J. Trump. In several pre-registered experiments, I examine three layers of polarization across multiple settings: I document the behavioral-, belief-, and norm-based mechanisms through which perceptions of interpersonal closeness, altruism, and cooperativeness are affected by polarization, both within and between political factions. To separate ingroup-love from outgroup-hate, the political setting is contrasted with a minimal group setting. I find strong heterogeneous effects: ingroup-love occurs in the perceptional domain (how close one feels towards others), whereas outgroup-hate occurs in the behavioral domain (how one helps/harms/cooperates with others). In addition, the pernicious outcomes of partisan identity also comport with the elicited social norms. Noteworthy, the rich experimental setting also allows me to examine the drivers of these behaviors, suggesting that the observed partisan rift might be not as forlorn as previously suggested: in the contexts studied here, the adverse behavioral impact of the resulting intergroup conflict can be attributed to one's grim expectations about the cooperativeness of the opposing faction, as opposed to one's actual unwillingness to cooperate with them

We study a model of the rise and fall of illiberal democracies. Voters value both liberty and security. In times of crisis, voters may prefer to elect an illiberal government that, by violating constitutional constraints, offers greater security but less liberty. However, violating these constraints allows the government to manipulate information, in turn reducing electoral accountability. We show how elements of liberal constitutions induce voters to elect illiberal governments that remain in power for inefficiently long—including forever. We derive insights into what makes constitutions stable against the rise of illiberal governments and extend the model to allow for illiberal governments to overcome checks and balances and become autocracies.

This paper develops a model of voters’ and politicians’ behavior based on the notion that voters focus disproportionately on and, hence, overweigh certain attributes of policies. We assume that policies have two attributes—resources devoted to two distinct issues (e.g., defense and education)—and that voters focus more on the attribute in which their options differ more. First, we consider exogenous policies and show that focusing polarizes the electorate. Second, we consider the endogenous supply of policies by politicians running for office and show that focusing leads to inefficiencies: voters that are more focused are more influential; distorted attention empowers social groups that are larger and more sensitive to changes on either issue; resources are channelled towards divisive issues. Finally, we show that augmenting classical models of electoral competition with focusing can contribute to explain puzzling stylized facts such as the inverse correlation between income inequality and redistribution.

Can politics intensify the role of religion and impact attitudes? The faith-based initiatives are a series of reforms in the USA that strengthened the ties between government and the religious community and made it more worthwhile to set up religious organizations. We exploit the different uptake of the initiatives across US states in a differences-in-differences setup. First, we find that the initiatives intensified religious attendance, beliefs, and increased the number of religious nonprofit organizations. In a back-of-the-envelope calculation, we show that the US rate of secularization would have been three times as rapid, had it not been for the faith-based initiatives. States did not differ in terms of changes in religiosity or potentially important confounders prior to the reforms. Next, we document that the initiatives increased skepticism towards homosexuals, modern gender roles, and science and strengthened preferences for crime reduction. The results point to politics as one explanation for the continued high average US religiosity and contribute to our understanding of how religion may impact outcomes.

We propose a theory of democratic backsliding where citizens retrospective assessment of politicians depend on expectations that are endogenous to the incumbent behavior. We show that democratic backsliding can occur even when most citizens and most politicians intrinsically value democracy. By challenging norms of democracy, incumbents can lower citizens’ expectations—only to partially back down and beat this lowered standard. As a result, gradual backsliding can actually enhances an incumbent’s popular support not despite but because citizens value democracy. This mechanism can only arise when, owing to programmatically weak parties, citizens are uncertain enough about leaders’ policy preferences. Conversely, mass polarization and citizens’ information can reduce the occurrence of backsliding, yet increase its severity.

We consider a dynamic election model in which a voter faces a trade off between the need to adapt the policy to a changing state and a desire for policy stability. In each period, the voter observes an imperfect signal about the state, and then delegates the policy making to one of two parties. Each party is more informed than the voter, but is ideologically biased relative to her preferences.

In the static model, the voter faces two delegation inefficiencies: 1. the elected party may inefficiently leave the policy in place, and 2. the elected party may inefficiently change the policy. Facing the tradeoff between these two inefficiencies, we show, the voter exhibits a status-quo bias: she tends to elect the party whose ideology is aligned with the status quo. In the dynamic model, expecting this electoral bias, the party in office becomes less responsive to the state in order to be reelected, and instead too frequently implements the policy it is ideologically aligned with. Hence, the electoral strategy of the voter, though optimal ex post, exacerbates both delegation distortions and leads to an apparent polarization of the parties.

Delegation distortions are largest when the need for delegation is the largest, that is, when the voter is least informed about the state. Delegation distortions also increase with parties' office motivation. In particular, even if parties are ideologically arbitrarily close, office motivation can induce them to behave in a strongly polarized way.

When a digital platform such as Google News selects personalized news for its user, will it select news that conforms to its user’s existing bias, thus creating an “echo chamber”? To answer this question, this paper studies a game between a click-maximizing platform and a user who tries to learn the true state of the world. This paper shows that, contrary to popular belief, it is optimal for the platform to select news that contradicts the user’s existing bias. This result stands in contrast with the bias of traditional media such as newspapers (Gentzkow and Shapiro, 2006; Suen, 2004) and this contrast is consistent with the recent empirical findings on online news consumption. This paper shows that it is optimal for a platform to send its user news that opposes her current view for two reasons. On the one hand, the user prefers opposing news because she expects to learn more about the state of the world from it, even if she expects it to be less credible than the news that agrees with her views. On the other hand, by sending surprising news, the platform challenges the user’s belief about the true state and increases her demand to click for more information.

Mass shootings have increased rapidly in the last two decades in the U.S. In this paper, we investigate the economic consequences of mass shootings. We find that shootings have negative effects on targeted counties' employment, earnings, and housing prices. Examining the mechanisms, we find that residents of targeted areas: (i) are more likely to report being unable to do their usual activities such as working, suggesting shootings lead to a decrease in productivity; (ii) have pessimistic views of financial and local business conditions. Further, we find that greater national media coverage of shootings exacerbates their local economic consequences

Economic growth needs a strong and well-functioning government. But a government too strong can dominate private firms, leading to a holdup problem that is especially severe in autocracies. This paper studies how to constrain officials in autocracies through personnel rules, with a special focus on rotation and performance evaluation. Through a game theoretic model, I show that rotation or performance evaluation alone actually makes holdup problems even worse. But it is exactly their combination that covers each other's weakness and solves holdup problems together. Frequently rotated and carefully evaluated, officials also develop few entrenched interests in existing firms. This helps avoid crony capitalism and encourages Schumpeterian "creative destruction", solving another key problem with government-assisted development. Thus, rotation and performance rewards resolve the acute tradeoff between commitment and flexibility, a feature rarely satisfied by other commitment devices. Firm-level panel data from China are consistent with the key predictions of the model.

It is difficult to know if racial discrepancies in police stop and search data are caused by racial bias or statistical discrimination. In part, this is due to uncertainty over the benchmark of unbiased police behavior: do officers aim to maximize arrests or to minimize crime? In this paper, I compare models of the two police objectives to data from U.S. cities. Empirical evidence is consistent with a model of arrest maximization and inconsistent with a model of crime minimization. These findings support the validity of existing tests for racial bias that rely on the assumption that police maximize arrests.

The public health crisis caused by the COVID-19 pandemic, coupled with the subsequent economic emergency and social turmoil, has pushed governments to substantially and swiftly increase spending. Because of the pressing nature of the crisis, public procurement rules and procedures have been relaxed in many places in order to expedite transactions. However, this may also create opportunities for corruption. Using contract-level information on public spending from Colombia’s e-procurement platform, and a differencein-differences identification strategy, we find that municipalities classified by a machine learning algorithm as traditionally more prone to corruption react to the pandemic-led spending surge by using a larger proportion of discretionary non-competitive contracts and increasing their average value. This is especially so in the case of contracts to procure crisisrelated goods and services. Our evidence suggests that large negative shocks that require fast and massive spending may increase corruption, thus at least partially offsetting the mitigating effects of this fiscal instrument.

Most of the innovation literature has found that distance is a barrier to the diffusion of knowledge, based on distances between citing and cited patents. In this paper, I investigate whether distance is a barrier to the spread of AI employment from regions with historical strength in AI scientific publications. I merging Burning Glass data for online job advertisements from 2007-2019 with data on AI scientific publications from 1950-2006 to form a panel of commuting zones. I find that commuting zones distant from AI publishing hotspots have slightly slower growth in AI employment than closer commuting zones, and I identify the number of publications a commuting zone must have to generate spillover effects.

Decision makers like regulators and bureaucrats, when called to evaluate or approve a reform proposed by an interest group or a legislator, suffer from a double asymmetric information, being uncertain about alignment and competence of the reform proposer as well as about the common value importance of the proposal and its consequences. We characterize the equilibria of this strategic interaction between reformers and decision makers as a function of complexity, and study the dynamics of endogenous complexity and stability of rules. We find that greater complexity may induce greater incentives to adopt reform proposals, which in turn can affect complexity. Complexification-simplification cycles can occur on the equilibrium path, and the most likely expected complexity in the long run may be higher when competence or alignment of reform proposers is lower. The results apply to regulation and organization, legislative politics and institutional design.

By filtering the information that reaches citizens, media coverage influences voter perceptions, government policy, and overall political accountability. A growing literature examines how political and economic forces shape media content at the newspaper level, but we have little evidence at the journalist level. We provide direct evidence of how individual journalists' incentives affect media content and tone. An online media firm in Kenya randomly switched some of their writers from a per-article piece rate to a pay-per-view (PPV) contract. Writers on the PPV contract produced---as intended---more popular articles, but they submitted substantially fewer articles on average than writers receiving a piece rate. The PPV contract had particularly strong effects on risk averse writers' participation---both at the intensive and extensive margins. This suggests that contract type can have important implications for labor force composition in the gig economy. While we see no impact of the PPV contract on clickbait or overall quality, the increased pageviews do come at a price: PPV writers sharply reduce their supply of local news and their political coverage uses more negative language.

Can religious beliefs, and specifically beliefs in the supernatural, influence a society's likelihood to wage war? Plagued with endogeneity concerns, this question has concerned historians, political scientists, religious scholars, anthropologists, and sociologists for millennia. We suggest an exogenous driver of supernatural beliefs that is neither related to preceding wars, the economy, political factors, or any other anthropoid characteristics: solar eclipses in the Middle Ages. We study interstate wars in 104 Western European states from 1400-1800, a time when the scientific understanding of solar eclipses was undeveloped and their occurrence was regularly considered a supernatural sign. Our findings suggest the likelihood of interstate war onset rose by 3.5 percentage points in a year in which a state's capital experienced a solar eclipse. We also find meaningful, positive effects when predicting war incidence, whereas the likelihood that an ongoing war ends decreases by one percentage point in solar eclipse years. All results prevail, and remain largely unaffected in magnitude, when accounting for time trends (linear and squared), latitude and longitude of the state's capital, state-fixed effects, decade-fixed effects, and state-year-specific time trends. Further, randomly generating solar eclipse observations in our sample of the Middle Ages does not generate estimates that come close to our actual findings. Placebo regressions show no statistically meaningful link between solar eclipses and interstate war between 1960 and 2018, when the scientific knowledge about solar eclipses was advanced sufficiently to understand their scientific causes. Overall, these results suggest a positive, causal relationship between supernatural beliefs and the propensity to wage war.

We study the process of modernization of a State through the lenses of the U.S. federal government, analyzing how its bureaucratic structure, selection and promotion policies changed from 1816 to 1921. We digitize archival records containing the roster of all U.S. federal employees over this period, with information on their position in the bureaucratic hierarchy, specific occupation and compensation. We complement this data with information on the date in which each position in the U.S. federal government transitioned from a patronage system to the civil service, collected through reading of archival documents. Using this unique dataset, we study the organization of a State over the different stages of its development and document how its organizational characteristics change during the process of modernization. We then study and measure the role of political influence across this period. The long time span allows us to investigate these patterns from the early years of the U.S. bureaucracy to the height of its patronage system, until its period of transition to a modern civil service.

Does armed conflict reduce trade even in non-combat areas through the destruction of inter-group social capital? We analyze Ukrainian trade transactions before and after the 2014 Russia-Ukraine conflict. In a difference-in-differences framework, we find that Ukrainian firms from districts with fewer ethnic Russians experienced a deeper decline in trade with Russia. This decline is economically significant, persistent, and explained by erosion of trust and the rise of local nationalism. Affected Ukrainian firms suffered a decrease in performance and diverted trade to other countries. Our results suggest that, through social effects, conflict can be economically damaging even away from combat areas.

In this paper, we study how policies limiting the spending capacity of local governments may lead to a reduction in corruption. We exploit the extension of one such policy, the Domestic Stability Pact (DSP), to Italian municipalities with less than 5,000 inhabitants that occurred in 2013. Using a ‘local Difference-in-Differences’ approach, we show that the extension of the DSP led to a substantial decrease in recorded corruption rates. This effect emerges only in areas in which the DSP put a binding cap on municipal capital expenditures, in line with the hypothesis that investments and procurement are naturally prone to corruptive phenomena. We also show that i) the reduction in corruption is linked to accountability incentives; ii) and it is not just a mechanical consequence of the decrease in investments, by pointing out evidence of an improvement in the corruption-proofness of public spending. We then estimate the impact of the extension of the DSP on local welfare, finding a null effect. Overall, our findings suggest that budget constrains might induce local governments to curb expenditures in a way that dampens their exposure to corruption without depressing local welfare.

We study how fear can affect the behavior of voters and politicians by looking at the Ebola scare that hit the U.S. a month before the 2014 midterm elections. Exploiting the timing and location of the four cases diagnosed in the U.S., we show that heightened concern about Ebola, as measured by online activity, led to a lower vote share for the Democrats in congressional and gubernatorial elections, as well as lower turnout, despite no evidence of a general anti-incumbent effect (including on President Obama’s approval ratings). We then show that politicians responded to the Ebola scare by mentioning the disease in connection with immigration, terrorism, and President Obama in newsletters, tweets and campaign ads. This response came only from Republicans, especially those facing competitive races, suggesting a strategic use of the issue in conjunction with topics perceived as favorable to them. Survey evidence suggests that voters responded with increasingly conservative attitudes on immigration but not on other ideologically-charged issues. Taken together, our findings indicate that emotional reactions associated with fear can have a strong electoral impact, that politicians perceive and act strategically in response to this, and that the process is mediated by issues that can be plausibly associated with the specific fear-triggering factor.

During periods of domestic turmoil, governments have an incentive to divert the attention of the population away from pressing domestic issues. I test whether such diversion occurs through strategic foreign interactions, using textual data retrieved from the GDELT database, for a globally representative sample of 195 countries, at the monthly level, over the years 1997-2017. I find robust evidence that governments resort to diversionary tactics in times of domestic turmoil and that such diversion takes the form of 'verbally' aggressive foreign interactions, typically targeted at ’weak’ countries and countries closely linked along genetic, religious, linguistic and geographic dimensions. Countries sharing a close trade relationship are unlikely to be victimized. These findings suggest that diversionary foreign policy is, in fact, systematically practised by governments as a strategic tool, and that such diversion is exercised in a manner that may not lead to large scale costs or risks of retaliation

In their highly influential paper, ‘Does Trade Cause Growth?’, Frankel and Romer estimate a trade equation to predict bilateral trade shares, which are in turn aggregated to construct an instrument for trade openness in income regressions. The Frankel-Romer approach has gained widespread popularity as a method to generate instruments for trade, foreign aid, FDI, immigration, knowledge diffusion, immigration diversity etc. from bilateral relationships. This research shows analytically and empirically that the Frankel-Romer instrument gives misleading results when fitted shares for zero and missing bilateral trade are omitted from the instrument set, because the exclusion restriction is violated. That is so because the instrument, due to the aggregation, captures the number of trading partners of a country as well as some of its observed trade. Our results imply that all the predicted bilateral trade relationships must be included in the instrument for trade openness to avoid mechanical endogeneity.

We conduct a survey and incentivized lab-in-the-field experimental tasks in Tirana, Albania. Our study was transformed from a migration study into a natural experiment by two large earthquakes that shook the Tirana area during our data collection period. These events provided a rare opportunity to gather evidence (including a pre-earthquake control) on the effect of natural disasters on time and risk preferences. We find unambiguous effects towards more risk aversion and impatience for affected individuals. Moreover, as it turns out, the second earthquake amplified the effect of the first one, suggesting that experiences cumulate in their influence on these preferences.

Individuals in low-income settings are often overly pessimistic about their own survival, suggesting that better knowledge about survival risks might encourage investments in health. This paper provides evidence from a randomized experiment that provided mature adults aged 45+ in Malawi with information about mortality risks. Treated individuals are less likely to engage in risky sexual practices one year after the intervention, and they increase other forward-looking behaviors such as investments in agriculture. Expectations of HIV+ people living longer, which makes the pool of potential partners riskier, are a primary driver of reduced sexual risk taking in response to the intervention.

German separation in 1949 into a communist East and a capitalist West and their reunification in 1990 are commonly described as a natural experiment to study the enduring effects of communism. We show in three steps that the populations in East and West Germany were far from being randomly selected treatment and control groups. First, the later border is already visible in many socio-economic characteristics in pre-World War II data. Second, World War II and the subsequent occupying forces affected East and West differently. Third, a selective fifth of the population fled from East to West Germany before the building of the Wall in 1961. In light of our findings, we propose a more cautious interpretation of the extensive literature on the enduring effects of communist systems on economic outcomes, political preferences, cultural traits, and gender roles.

To what extent can heroes coordinate and legitimize otherwise strongly- proscribed and potentially repugnant political behavior? In this paper, we exploit the purposefully arbitrary Noria rotation of French regiments to measure the legitimizing effects of heroic human capital, gleaned through exposure to the pivotal Battle of Verdun under General Philippe Pétain in 1916. We wed this with a unique newly declassified dataset of 97,242 individual collaborators with the Nazis collected by French army intelligence in 1945 to show that, during the Pétain-led Vichy regime (1940-44), municipalities that raised troops that served under Pétain at Verdun later housed more collaborators with the Nazis than otherwise similar municipalities. Residents of these municipalities were 5 to 10% more likely to join Fascist political parties or paramilitary groups that conducted the internal repression of the regime against Jews and resistants, and to directly join German military units. We interpret these results as reflecting the role that Verdun played in generating both credentials for leadership and organizational capacity that legitimized otherwise proscribed values, forging political identities that proved durable in explaining the Left-Right divide in France throughout much of the post-war period, and was particularly salient in times of social and political crisis.

The seminar will present the results from new experimental evaluations of unconventional policies to re-employ unemployed workers. These policies are particularly valuable at a time when face-to-face employment services are difficult to provide (e.g., during a pandemic). A first set of experiments was implemented in Italy to evaluate two types of financial incentives for re-employment of job seekers at risk of long-term unemployment: (i) a re-employment voucher that incentivizes a specialized third party to match the job seeker with an employer; and (ii) a re-employment bonus that incentivizes the job seeker directly. Each policy turns out to be more effective than an alternative, conventional job search assistance program based on face-to-face interactions, and we suggest the possibility of a dual voucher--bonus system based on self-selection. A second experiment was implemented in Australia to evaluate a website that provides job search assistance online, through editable resume and cover letter templates as well as tips on how to look and apply for jobs. The intervention increased job-finding rates, particularly among job seekers aged 35--50, with larger effects for women in this age group. The quality of job matches improved too.