#Econhist papers I (mostly) admire, Aug 2023

On African marriage records, colonial trade, bison in America, lessons from the Dust Bowl, and more...

A new feature for paid subscribers launches this month. I plan to review, once a month, some of the most interesting new research in economic history. I hope to avoid the economics (or scientific) jargon that often makes scholarly papers inaccessible even to scientists in neighbouring fields. My preferences will, as always, be biased towards African economic history, but I suspect that the research I review will tend to cover many geographies, time periods and even fields outside economic history, as is the case in this first post. This series will be paywalled for one month before opening up to free subscribers.

The title of the series is, of course, adapted from one of my favourite podcasts: Steve Levitt interviews people he mostly admires. Make sure to tune in to PIMA wherever you get your podcasts.

Now, on to the latest in economic history…

[A paid subscription not only gives you unlimited access to all of this content but also supports the continued production of these articles. Please consider subscribing.]

1. What we learn from marriage registers in Africa

Felix Meier zu Selhausen and Jacob Weisdorf have a new working paper on gender inequality in colonial Africa.

Five-point summary:

1. The study analysed data from 30,000 marriage registers from six major cities in British Africa, exploring the effects of mission schooling and colonial influence on African men and women’s education and occupations between 1860 and 1970.

2. Although both men and women had access to mission education, the benefits were uneven: men were more likely to gain access to formal jobs while women were often relegated to homemaking or informal work.

3. A significant difference was noticed regionally: West African women, due to their precolonial traditions of economic independence, were more likely to engage in informal income-generating activities, unlike East African women, who were mostly confined to domestic tasks.

4. Over the colonial period, the trend of women being segregated from formal work decreased, particularly in West Africa. This shift was linked to the Africanisation and feminisation of the British public service and the influence of Western women’s liberalisation movement post-World War II.

5. Women’s opportunities were heavily influenced by their social background and their fathers’ roles within colonial society. Daughters of men working in formal colonial roles had better chances of attending mission schools and entering the formal economy, contributing to less gender inequality in their subsequent marriages.

My takeaway:

In the absence of detailed census records, economic historians have to be creative in finding good sources to work with. Felix’s remarkable efforts to collect more than 30,000 marriage records expose the complex ways colonial-era missionaries affected the gender dynamics within African society, influencing educational opportunities, career prospects, and social advancement for both men and women across various regions in British Africa.

2. Bison and America’s reversal of fortunes

I met Rob Gillezeau recently for lunch in Stellenbosch, and as a result, I discovered this amazing paper published in The Review of Economic Studies.

Five-point summary:

1. The North American bison, a primary resource for the Native Americans, reached near-extinction at the end of the 19th century, causing a significant change in the living conditions of Indigenous nations that relied on them for food, clothing, shelter, and tools.

2. The decline of the bison resulted in an immediate decline in the material well-being of bison-reliant nations, persisting into the present and contributing to the relative poverty of Indigenous nations in North America today.

3. Data shows the negative health effects of the loss of the bison, with evidence of higher rates of child mortality, skewed sex ratios, and a 2.5 cm decline in height in bison-reliant nations.

4. Instead of adjusting over time, these nations experienced lower levels of income per capita, leading to a different economic trajectory for bison nations, including changes in population, income, and credit markets.

5. This historical event highlights the persistent impacts of economic shocks, the importance of financial intermediation in assisting economic recovery, and the counter-narrative about the colonisation of North America and the disruption of Indigenous resources and industries.

My takeaway:

This story of colonial encroachment and dispossession is, of course, not unique to North America. In some cases, the shock of European arrival was more extreme (consider the Khoe of the Cape Colony), while in other places, native inhabitants managed to not only maintain some independence but would actually accede to the same living standards as the European colonisers (see New Zealand). Explaining these diverse experiences remain important.

3. To live longer, choose your ancestors well

A new Nature Communications paper shows that your lifespan and health span are strongly correlated to that of your parents and grandparents.

Five-point summary:

1. Human longevity, or the ability to live to an old age, has been found to be heritable and tends to cluster in certain families. Members of these families not only live longer but also appear to age more healthily compared to the general population.

2. An extensive study conducted over 26 years used both family-based and population-based data to examine the development of diseases from mid-life onwards. The study revealed that members of these long-lived families exhibit a delayed onset of disease, multimorbidity (the occurrence of multiple chronic diseases), and medication use, thereby extending their health span by up to a decade.

3. The study employed a quantitative measure known as the Longevity Relatives Count (LRC) score to quantify familial longevity. It was found that an increasing number of long-lived ancestors, as measured by the LRC score, correlates with an increasing delay in disease onset and lower medication use.

4. Individuals with a higher LRC score (≥0.60) had a better MetaboHealth score, indicating superior immune and metabolic health, as well as a lower risk of mortality over a five-year period. For each 10% increase in LRC score, there was a 5% reduction in the annual risk of developing an age-related disease in the Dutch cohort (LLS), and a 6% reduction in the Swedish cohort (SEDD). This risk reduction could reach up to 50% and 60%, respectively, if all ancestors were long-lived.

5. The findings highlight the importance of integrating multi-generational ancestral mortality data into studies. This approach could enable researchers to elucidate mechanisms that protect against multimorbidity and promote healthspan already in mid-life and uncover the complex interplay between genetic and environmental factors. In the future, these findings could be applied across studies to enhance the comparative nature of longevity studies and facilitate the discovery of novel genetic variants and mechanisms promoting healthy ageing.

My takeaway:

Genealogical data is becoming increasingly useful to test not only the persistence of health outcomes across generations, as this study shows, but also incomes. I think this is one of the most exciting fields in economics. (Many years ago, I also contributed to a paper about the multigenerational persistence of longevity.)

4. How to measure inequality in data-poor settings

My colleagues Dieter von Fintel, Calumet Links and Erik Green have a new working paper on social tables.

Five-point summary:

1. The study of long-term historical inequality has expanded to include previously overlooked periods and societies, especially in the global South. This expansion has been facilitated by the resurgence of the social tables method, which uses archival records to reconstruct income and wealth distributions where microdata is not available.

2. The researchers collected a new dataset of 108 historical social tables spanning over 1,000 years to examine long-term historical inequality, particularly in the global South, using a method known as the social tables approach.

3. The study found no systematic relationship between inequality estimates and the number of classes chosen or the size of the top class. However, when bottom classes bundled together even small variations in income or wealth, it could introduce a downward bias to the inequality estimates.

4. The researchers identified that the size of the bottom class in social tables significantly impacts inequality estimates in populations smaller than two million. This highlights a previously overlooked aspect in inequality estimation: the role of large bottom classes.

5. The study makes a call for better statistical methods, exploration of new archival sources, and the need for agreement on conventions for choosing the size of social table classes. This is to mitigate bias, ensure estimates from various settings are harmonised, and facilitate the production of comparable estimates that can provide a global view of historical inequality.

My takeaway:

Social tables are useful in the absence of microdata. But they may introduce biases, as this study shows, and make inequality estimates difficult to compare. Just as there is agreement on how to measure standards of living, so economic historians need to find agreement on what the best standards are for social tables. This paper is a good start for that discussion.

5. What we learn from eighteenth-century American ports

Gothenberg University economic historian Jeremy Land has just published an excellent book on eighteenth-century colonial trade and its role in the American Revolution.

Five-point summary:

1. This book reassesses the role of Boston, New York, and Philadelphia in colonial American trade and their influence on the global economy prior to America’s independence. It proposes that these cities were the primary ports within a large, interconnected port complex, serving as key nodes in British imperial and Atlantic trade.

2. It examines these cities’ trade practices within and beyond the British Empire, arguing that their economic orientation was significantly more global than previously considered. This global reach often conflicted with the British Empire’s mercantilist policies, leading to increasing resistance from colonial merchants.

3. The book questions traditional categorisations of these three cities into separate regional zones and emphasises their similarities, competition, and cooperation, facilitating a clearer understanding of their relationship with the expanding global economy of the eighteenth century.

4. It provides a detailed analysis of the British Empire’s inability to effectively enforce its trade regulations and its limited fiscal support for its colonies. This lack of ‘imperial capacity’ inadvertently allowed colonial merchants to cultivate alternative trade networks and practices, increasing their access to global markets.

5. The research concludes with an exploration of how these merchant networks developed amidst the absence of British state power, suggesting that this relative independence played a significant role in America’s subsequent journey towards independence and its economic development.

My takeaway:

This book is based on Land’s WEHC 2022-prizewinning dissertation. It is a must-have for anyone working on the eighteenth-century Atlantic economy.

6. Does technological progress inevitably lead to deskilling?

You would expect the stellar team of Jeremy Atack, Robert A. Margo and Paul Rhode to ask important, creative and relevant questions. This NBER working paper on deskilling in late-nineteenth-century American manufacturing is no different.

1. The traditional understanding among historians is that the rise of the steam-powered factory resulted in ‘deskilling’ where machinery operated by semi-skilled workers took over tasks previously performed by skilled artisans using simple tools. This was supported by data showing that between 1850 and 1910, the share of skilled blue-collar workers declined by 17 percentage points, indicating a shift towards less-skilled labour.

2. This study finds compelling evidence of the ‘deskilling’ trend in the late nineteenth century, with 36 per cent of operations requiring less skilled labour in machine production than in hand production. This indicates a substantial shift from artisan craftsmanship to semi-skilled and unskilled labour in mechanised factories.

3. However, contrary to common perception, the mechanisation of factories contributed to only a fraction of this deskilling trend. Specifically, mechanisation accounted for just 7 per cent of the mean value of the De-Skill measure, suggesting that other factors played more significant roles in the deskilling process.

4. Notably, the division of labour emerged as a more potent force driving deskilling. Regression analysis reveals a strong correlation between the increased division of labour and higher levels of deskilling, indicating that the fragmentation of production tasks among separate workers significantly reduced the demand for skilled artisans.

5. The increase in the division of labour was closely tied to the growth of establishment size and the expansion of the market, both of which were influenced by the transportation revolution. The larger the scale of operation and the broader the market access, the more advantageous it became to divide labour, leading to more specialised workers and less reliance on skilled artisans.

My takeaway:

There is much concern about how artificial intelligence may cause the ‘deskilling’ of the workforce. Historical lessons, like this late nineteenth-century example, point to the multiple ways new technologies may affect the workforce.

7. How people respond to environmental change

Rick Hornbeck has a new paper in the Journal of Economic History on the migration response to environmental changes.

Five-point summary:

1. The Dust Bowl of the 1930s resulted in significant land erosion, causing a reduction in agricultural land values, declines in population, and triggering large-scale migration, particularly from the most-eroded counties.

2. Migration rates were higher from more-eroded counties, and these migrants moved further and to more scattered locations, suggesting an unusual and potentially less-planned response to the environmental crisis.

3. Migrants from more-eroded counties had fewer years of education than non-migrants and other migrants. These migrants were typically younger, more likely to be male, and more likely to have lived on a farm in 1935.

4. The New Deal program spending had mixed effects. While greater agricultural adjustment administration spending had a more negative effect on male incomes, public works spending moderately mitigated the impact on male incomes, but no substantial impact was found on female incomes or due to relief spending.

5. The Dust Bowl offers a unique historical case study of large-scale migration in response to environmental collapse, providing insights into migration as an adaptation strategy, and serving as a potential blueprint for understanding future climate-change-induced migrations.

My takeaway:

With concerns about global climate change intensifying, these historical episodes help to understand the likely consequences of environmental-induced migration.

8. Our experiences are not sufficient to understand the world

Finally, Max Roser, founder of Our World in Data (of which I am a frequent user), made the case for global statistics in a recent blog post.

Five-point summary:

1. Personal experience alone is insufficient to truly understand the world due to the sheer number of people and the diversity of their experiences. Even the most social individuals can only directly interact with a tiny fraction of the world’s population.

2. The people we know tend to be similar to us and therefore do not fully represent the global population or even our own country.

3. We rely on various media outlets to learn about people and events beyond our personal experience, but these sources often focus on unusual events or exceptional individuals, leaving us with a fragmented view of the world.

4. Global statistics are crucial for a comprehensive understanding of the world, as they can cover everyone and provide information about regular and everyday occurrences often overlooked by the media.

5. While statistics offer a broader view, they also have limitations and are not always perfect. It is important to critically analyse the data and understand its limitations to gain a comprehensive view of the world.

My takeaway:

This is true even at the regional or local level. Our experience of our country, our region, and even our town or neighbourhood is but one of many experiences. I have to remind myself about this often as a columnist. That is why statistical information that covers the full population is so vital to inform our decisions and policies.