| Abstract |
1. Trade Network Centrality and Invoicing Currency
This paper studies how global trade network structure shapes countries' invoicing currency choices. Using panel data on trade invoicing and network centrality, we show that countries use more of their home currency for import invoicing as their structural centrality increases. In contrast, home-currency usage in export invoicing exhibits a non-linear relationship with network position, emerging only at very high levels of structural centrality. These findings highlight the role of asymmetric trade networks in sustaining dominant currencies and show that invoicing responds to long-run structural network position rather than short-run trade intensity fluctuations.
2. Asset bubbles in a two-sector overlapping generation model
This article studies the link between asset bubbles, outputs, relative price of goods and physical capital in a two-sector (consumption and investment) two-period overlapping generations model. In this framework, agents save through physical capital investment and an asset bubble with no fundamental value. The formation of the bubble diverts savings away from capital accumulation, causing investment goods production to decline. Simultaneously, the bubble raises perceived wealth, increasing consumption demand and boosting output in the consumption goods sector. As a result, even though a crowd-out effect of capital stock may exist, aggregate GDP can rise due to the consumption production expansion.
3. Saving by Shopping: A Theory of Time Allocation and Evidence from Scanner Data
How often should households go shopping? This paper studies endogenous shopping behavior within a household production framework using novel retailer scanner data. The data provide transaction-level information linked to individual shoppers, allowing us to estimate the savings associated with more frequent shopping. We develop a model in which frequent shopping increases the likelihood of purchasing items on sale, while households optimally allocate time across shopping, labor, and leisure. Using this framework, we jointly estimate shopping costs and the elasticity of substitution between time and market goods. The estimated elasticity is substantially lower than one and values commonly assumed in macroeconomic models. Welfare analysis shows that changes in shopping-specific costs have modest effects, whereas increases in time costs that reduce overall available time lead to sizable welfare losses.
4. Good Lord, Bad Democrat? Institutional Quality under Autocracy and Long-Run Political Persistence in Democracy
Do well-functioning institutions under nondemocratic regimes foster long-run democratic capacity? This paper examines the persistent political and administrative consequences of competent autocratic governance, using Tokugawa-era Japan (1600–1868) as a historical laboratory. While domains shared a common institutional framework, they varied substantially in administrative effectiveness, allowing comparison of institutional performance holding regime type constant.
I construct novel domain-level measures of governance capacity—riot frequency, the density of Dutch-trained scholars, and market discounts on domain currencies—and identify exogenous variation using two sequential shocks. The 1588 Katanagari sword hunt fixed the number of sword-bearing retainers per lord, while the 1600 Battle of Sekigahara reassigned territories and populations without redistributing retainers. This generated plausibly exogenous variation in administrative density across domains. I trace the long-run effects of this variation to the interwar period, focusing on a large-scale rural development program (1931–1937) that offered guaranteed funding to municipalities conditional on submitting locally coordinated reconstruction plans. Despite identical incentives, many villages failed to apply. At the municipal level (N > 8500), I show that regions with more competent feudal governance were significantly less likely to produce successful bottom-up plans and exhibited lower civic initiative, alongside greater support for authoritarian parties in the 1937 election. The findings reveal a paradox: effective autocratic governance strengthened state capacity but weakened local collective action, suggesting a long-run tradeoff between administrative competence and civic engagement. |