This study examines the effect of media use on media trust and persuasion using a largescale randomized field experiment, which was conducted in collaboration with the nation's most trusted media outlet. By randomly increasing the capacity for viewing its TV programs, Kitamura and Kuroda found that this treatment increased support for government policies by increasing program viewing time, which is, as they demonstrate, biased in favor of the government. Furthermore, the researchers determined that the effect is driven mostly by those who trusted the outlet more than other broadcasters and that their levels of trust in the outlet were even increased by their treatment, which Kitamura and Kuroda call endogenous persuasion. By contrast, they did not discover heterogeneous effects with respect to political preferences. To better understand the mechanism underlying these findings, the researchers developed a model of endogenous persuasion.
Yamaguchi, Braguinsky, and Nakajima examine patents granted to university-educated scientists and engineers in Japan during its early industrialization. Their data encompass the census of all inventors with university degrees in science and engineering (S&E inventors) from the inception of higher S&E education in Japan and until the 1920 graduation cohorts and all patents from 1885-1940. The researchers find that patents by S&E inventors concentrated in fields related to new industries that took off later in the sample, so that knowledge accumulation was a key to industry growth; and that academic achievement, especially at the university level, predicted both sorting into invention among all S&E graduates and the number and quality of patents by those who became inventors. A deeper dive into the data reveals that those who chose careers in research were generally the most prolific inventors, closely followed by those who chose private companies. Accumulated work experience also increased S&E inventors' propensity to patent, especially for those at the top tail of the university academic achievement and latercohort graduates. Thus, Yamaguchi, Braguinsky, and Nakajima observe strong complementarity between work experience and absorption of specialized higher university S&E education and between work experience and graduating from a later cohort. The researchers conjecture those later graduates may have benefited from work experience in organizations that already employed inventors who graduated earlier.
Whether transaction costs to assemble or split land can persistently hinder efficient land use remains unknown in the context of cities. Constructing a 100 m*100 m-cell-level dataset of central Tokyo from the 19th-century pre-modern era to the 21st-century skyscraper era, Nakajima, Teshima, and Yamasaki study how initial lot fragmentation has affected urban development by exploiting the plausibly exogenous supply shock of large lots in 1868, the release of local lords' estates (daimyo yashiki) scattered throughout central Tokyo. Using ordinary least squares and a regression discontinuity design, they find that cells previously used as local lords' estates have larger lots today, implying that lot size persists by transaction costs. The researchers also find positive effects on land use and activities, that is, more taller buildings requiring large footprints and higher land prices and labor productivity of firms there, implying lot size premia due to assembly frictions. Nakajima, Teshima, and Yamasaki find these effects mainly in the core area, suggesting these areas have high transaction costs dominating potential benefits of assembly to construct skyscrapers there. In addition, the effect of lot size on land price turned from negative to positive after the rise of skyscrapers, suggesting greater importance of assembly friction due to the agglomeration economies that skyscrapers generate. Overall, transaction costs to assemble land in growing urban cores dominate the potential large benefits of assembly, which persistently hinder economic activities.
Do electoral institutions affect the degree to which female legislators address women's interests in legislative processes? While the growing literature has examined whether increased women's representation causally affects policy outcomes, whether electoral institutions mediate the effect is less known. To fill the gap, this study tests whether proportional representation (PR) encourages female representatives to address women-specific interests more than a single-member district (SMD) does. To elicit the causal impact of electoral institutions, Okuyama and Sudo leverage the unique "best loser" provision of the mixed electoral system in the Japanese House of Representatives elections, where a marginal candidate may win an SMD seat or PR seat by chance. To fully account for the complex structure of the mixed electoral system, they apply the simulation-based regression discontinuity design. Across different legislative activities, the researchers consistently find a significant effect of holding a PR seat: female PR representatives more frequently affiliate with women-related committees, submit question memorandums on women's issues, and endorse petitions regarding women's interests than their male counterparts, but significantly less so when they stand as SMD representatives. The institutional effect likely arises because a SMD representative has higher incentives to address issues both male and female voters care about. Such a vote-seeking strategy is not necessarily compatible with representing women-specific interests. Meanwhile, a PR representative earns their party's reputation from female voters by addressing women-specific interests. Overall, their results suggest that electoral institutions do affect the relationship between women's descriptive representation and their policy consequences. More broadly, their findings bring forward the research agenda in political economics to better understand the political institutions and policy choices and, in particular, underscore the importance of institutional environments in leveraging diverse voices in policymaking.
Past studies of Artificial Intelligence (AI) on employment examine its impacts across occupations and implicitly assume that all workers in the occupation are uniformly affected by AI, ignoring the substantial within-occupation skill heterogeneity. Kanazawa, Kawaguchi, Shigeoka, and Watanabe study the impact of a demand-forecasting AI on productivity in the context of taxi drivers. The researchers find that AI improves drivers’ productivity by shortening the search time for customers by 5%, on average. Importantly, the productivity gain is entirely accrued by low-skilled drivers, whereas the corresponding gain for high-skilled drivers is nearly zero, suggesting that AI is a substitute for skill, at least in this context.
Cartels participating in procurement auctions frequently use bid rotation or prioritize incumbents to allocate contracts. However, establishing a link between observed allocation patterns and firm conduct has been difficult: there are cost-based competitive explanations for such patterns. Chassang, Kawai, Nakabayashi, and Ortner show that by focusing on auctions in which the winning and losing bids are very close, it is possible to distinguish allocation patterns reflecting cost differences across firms from patterns reflecting non-competitive environments. Chassang, Kawai, Nakabayashi, and Ortner apply their tests to two datasets: the sample of Ohio milk auctions studied in Porter and Zona (1999), and a sample of municipal procurement auctions from Japan.
This paper analyzes individuals' medical expenditure risks over the life-cycle and roles of the national health insurance using nationwide administrative data of health insurance claims (NDB) in Japan. Health shocks are highly persistent and distribution of life-time expenditures varies by the assumed order of persistence. Fukai, Ichimura, Kitao, and Mikoshiba build a structural life-cycle model of males and females and single and married households with different skill levels and quantify economic and welfare effects of medical expenditure risks. The national health insurance characterized by age-dependent copay rates and progressive out-of-pocket ceilings protects households from expenditure risks well and affects their life-cycle saving substantially. Responses to a reform to reduce insurance benefits are heterogeneous: high-income households turn to self-insurance and raise savings, and low-income households, including low-skilled single females, reduce savings and consumption and many of them turn recipients of welfare transfers. Fukai, Ichimura, Kitao, and Mikoshiba also show that effects of health insurance reform depend on generosity of other welfare programs and do so differently across households.
Using smartphone data for Japan, Miyauchi, Nakajima, and Redding show that non-commuting trips are frequent, more localized than commuting trips, strongly related to the availability of nontraded services, and occur along trip chains. Guided by these empirical findings, they develop a quantitative urban model that incorporates travel to work and travel to consume non-traded services. The researchers use the gravity equation predictions of the model to estimate theoretically-consistent measures of travel access. Miyauchi, Nakajima, and Redding show that consumption access makes a substantial contribution to the observed variation in residents and land prices and the observed impact of the opening of a new subway line.
This paper was distributed as Working Paper 28497, where an updated version may be available.
Large-scale household inventory buildups occurred in Japan five times over the last decade, including those triggered by the Tohoku earthquake in 2011, the spread of COVID19 infections in 2020, and the consumption tax hikes in 2014 and 2019. Each of these episodes was accompanied by considerable swings in GDP, suggesting that fluctuations in household inventories are one of the sources of macroeconomic fluctuations in Japan. In this paper, Ueda, Watanabe, and Watanabe focus on changes in household inventories associated with temporary sales and propose a methodology to estimate changes in household inventories at the product level using retail scanner data. They construct a simple model on household stockpiling and derive equations for the relationships between the quantity consumed and the quantity purchased and between consumption and purchase prices. The researchers then use these relationships to make inferences about quantities consumed, consumption prices, and inventories. Next, Ueda, Watanabe, and Watanabe test the validity of this methodology by calculating price indices and check whether the intertemporal substitution bias they find in the price indices is consistent with theoretical predictions. The researchers empirically show that there exists a large bias in the Laspeyres, Paasche, and Törnqvist price indices, which is smaller at lower frequencies but non-trivial even at a quarterly frequency and that intertemporal substitution bias disappears for a particular type of price index if Ueda, Watanabe, and Watanabe switch from purchase-based data to consumption-based data.