Abstrac: This paper examines a previously undocumented historical asset bubble: the Office Mania. Leveraging a novel dataset of 1,910 office sales across eight Castilian cities between 1543 and 1714, and applying right-tailed unit root tests, we identify explosive price behavior starting in 1592. We demonstrate that this market satisfied Quinn and Turner’s “bubble triangle” framework (marketability, credit, and speculation). The boom was sparked by an investor flight to safety. As sovereign debt (juros) lost credibility during the 1590s fiscal crises, investors substituted offices for bonds. Consistent with rational-bubble models, this case illustrates how safe-asset shortages generate destabilizing, pre-modern financial manias.

Abstract: This paper analyses market integration in Spain during the 18th and 19th centuries, leveraging a newly hand-collected database of monthly wheat prices from 1725 to 1905 across several Spanish cities. Employing advanced econometric techniques, such as structural break test (Bai & Perron, 1998) and threshold error correction model (Hansen & Seo, 2002), we examine the evolution of the Law of One Price and market efficiency in Spain on the long-term. Additionally, panel data models are applied to assess the ultimate impact of trade policies and technological advancements on market integration.  Our findings reveal that trade liberalization policies initiated in 1756 promoted integration between Spain’s coastal regions and international markets, while simultaneously leading to greater economic isolation for inland (net-producer) cities. However, this trend reversed in the 19th century, as protectionist policies improved connectivity between inland regions and peripheral areas, though without yielding any discernible efficiency gain. The introduction of railways and telegraph systems in the 1860s marked a pivotal moment, enhancing market integration and substantially increasing market efficiency. This study provides new insights into the historical evolution of market dynamics and the complex interplay between policy, technology, and economic geography in the long-term.

Abstract: Homoploutia, the overlap between top labour and top capital incomes, is often seen as a hallmark of American capitalism since the 1980s, following decades in which it was supposedly almost absent. We revisit this view using a unique source: the 1936 Income Tax Project, for which the Roosevelt administration tabulated every federal income tax return by total income and by combination of income sources. The original returns were destroyed after the Second World War. We therefore reconstruct synthetic microdata for returns above $2,500 (roughly the top 5%), using a weighted bracket imputation consistent with the project’s tabulations, the Statistics of Income and the national accounts. Combined with tax and census-based data for later decades, this yields a series of homoploutia for 1936–2006. Homoploutia stood at 24.2 per cent in 1936, comparable to its levels in 1970 and 1990, and peaked at 38.1 per cent in 1950, its highest level of the century. Earlier survey-based estimates miss this pattern because they under-represent top earners before 1983 and do not separate wages from self-employment income. What changed over time was the composition of homoploutia. In 1936 it rested on salaried earners with capital income, while about half of top capital earners still reported no wages. Since 1950, “traditional” wage-plus-capital homoploutia has been broadly stable. Higher corporate and personal tax rates during and after the war encouraged the creation of partnerships, and proprietors’ income became the main driver of fluctuations in homoploutia. Corporate profits reinforced it, while growth in employee compensation dampened it. Homoploutia was the norm, not the exception, in twentieth-century America.

Abstract: Spain recorded its largest twentieth-century declines in mortality and gains in life expectancy during the 1940s and 1950s, precisely when output, real wages, public health spending, and food consumption were contracting sharply. This “demographic miracle” rests heavily on official infant mortality statistics. Before 1960, these statistics placed Spain among the European countries with the lowest day-0 and neonatal mortality, yet among those with the highest postneonatal and child (ages 1–4) mortality. We argue that this implausible combination reflects systematic under-registration of very early deaths, rooted in weak statistical instruments and a legal definition of birth that required 24 hours of survival. We exploit the 1957 Civil Registry Law and the new statistical bulletins introduced in 1960 as a quasi-natural experiment. Using an inverse synthetic control method with a donor pool of European countries at a comparable stage of the demographic transition, we estimate that official day-0 and neonatal mortality rates were biased downward by at least 42 and 22 per cent, respectively, before the reform. We then reconstruct neonatal, day-0 and infant mortality for 1925–1974 using different methodologies. The corrected series show that early-infant mortality fell more slowly than officially reported in the 1940s and early 1950s and faster thereafter. Mortality improvements under early Francoism remain substantial, but the sharpest gains begin only in the mid-1950s. Because they assume complete registration of postneonatal deaths, our estimates are conservative lower bounds.

Abstract: Violence is often included among the determinants of foreign direct investment (FDI), but existing studies rarely address reverse causality. We use the Mexican Drug War as a quasi-natural experiment. Felipe Calderón’s unexpected victory in the 2006 presidential election led to federal Joint Operations against drug cartels. These were rolled out across Mexican states from 2007 and were followed by a sharp rise in homicides. Their timing and location were uncorrelated with states’ pre-existing homicide rates, FDI, income or corruption levels. We combine state-level data for 1999–2015 with bilateral FDI flows from 48 origin countries and estimate the effect of violence using a structural gravity model (PPML with high-dimensional fixed effects), instrumental-variable GMM and the synthetic control method. Violence substantially reduces FDI: an increase of 20 homicides per 100,000 inhabitants lowers inflows by about 20 per cent. Effects vary widely across states. Synthetic controls show cumulative losses of up to 79 per cent (Durango), close to US$20 billion across eight treated states. Financial-sector FDI fell even in states where total inflows held up. To explore the mechanisms, we draw on the 2019 Economic Census and the National Survey of Business Victimization (ENVE). Both show that the direct costs of crime and of protection vary considerably by sector and firm size and are highest for service firms. Our findings suggest that the militarized strategy against drug trafficking carried large economic costs on top of its human toll.