City of My Dreams: Social Mobility and Growth in Stockholm since the Industrial Revolution
Are individuals’ economic, health, and social outcomes determined by where in a city they grow up? Are such (dis)advantages of place transmitted across generations? Can policy makers alter urban environments to produce better outcomes for disadvantaged children? This project develops a unique individual-level database of all inhabitants in Stockholm 1878-1926 linked to the building where they live. We use this data to analyze how economic opportunity, inequality, and segregation evolved as Stockholm was transformed into a modern metropolis around the turn of the last century. In particular, we study the returns to migration from moving to Stockholm and how different neighborhoods within the city shaped children’s economic and social outcomes. To identify causal effects, we exploit several natural experiments that induce both in-migration to Stockholm and within-city migration of families across neighborhoods. Moreover, we study how a range of policy interventions reshaped Stockholm and how changes in education, health infrastructure, and transportation affected individuals and the neighborhoods where they live. Our uniquely rich data allows us to study how urban environments shape the outcomes of several generations, as well as the historical origins of spatial inequality and segregation in Stockholm today.
Final report
City of My Dreams: Social Mobility and Growth in Stockholm since the Industrial Revolution
During the Industrial Revolution, Stockholm evolved from a small capital city with 50,000 inhabitants to a modern metropolis with several hundred thousand inhabitants. The city grew largely through the migration of both women and men from the countryside. This project uses a newly digitized and geolocated database covering the entire population of Stockholm 1878-1926 (the Roteman database) combined with, among others, the 1880-1930 censuses to investigate the effect of the move to Stockholm on migrants and how residential and childhood neighborhoods within the city affected the future life outcomes of Stockholmers.
Background and summary of the project
1. Purpose
The purpose of the project was to investigate how the emergence of large cities during the industrial revolution affected individuals' labor market outcomes, and to study how growing up in cities affects the local population in terms of labor market, family formation, health, and political participation.
2. Development of the project
The project has resulted in four studies. The first documents that large cities were an early contributor to women's entry into the labor market. Already around 1910, women's labor force participation was much higher in the largest cities in the US, UK and Sweden compared to rural areas and smaller cities. The study analyzes the issue further by examining the labor force participation of female migrants to Stockholm compared to their sisters migrating to other locations. The study also examines the effect on marriage, fertility, health, and political participation. We also examine how migration to Stockholm during industrialization affects intergenerational mobility, i.e. the extent to which children's income correlates with their fathers.
The second study examines the effect of social networks on whether women arriving in Stockholm start working. The study compares women who arrive in the same year and in the same area of Stockholm, but find housing in different buildings. It then examines how the composition of neighbors affects the likelihood of working, with a particular focus on the number of working women and men in the building. We also examine whether individuals seek out specific occupations held by their neighbors.
The third study examines the effect of childhood environment in Stockholm using detailed data on housing and neighbors during the first 16 years of life. We exploit the fact that cities were historically segregated vertically (within buildings) rather than horizontally (between buildings) as today. This meant that many poor children grew up in buildings with highly educated neighbors. We study the extent to which such potential interactions affected the life outcomes of boys and girls.
The fourth study examines the impact of Stockholm's expansion of water pipes that brought clean water to connected buildings. The study examines how access to clean water affected child mortality. Through access to individual data, we can also distinguish effect sizes based on household characteristics such as socioeconomic status. The study is almost ready to be submitted to a journal.
3. Key findings
Study one shows that large cities were extremely important for women's early entry into the formal labor market. Unlike men, who had equally high employment rates regardless of the population size of their place of residence, it was mainly large cities that gave women this opportunity. We also find that working women in large cities refrained from marrying and having children, which was probably a requirement to be able to work to a greater extent. Furthermore, we find that Stockholm migrants exhibit very high social mobility, with the correlation between father's and child's income being significantly lower than among individuals who do not move. This shows that urban migration largely broke the link between parents' and children's life outcomes.
The second study finds that social networks were important contributors to women's labor force participation. We find that women whose first residence was in a building with more working women are also more likely to work, both during their first year in Stockholm and after 5 years in the city.
The third study shows that the neighborhood has an important impact on individuals' future working life. In this study, we find that children of parents with blue-collar occupations are positively affected by having more adult neighbors with high occupational complexity. In particular, we find that this effect is mainly driven by having neighbors of the same gender as the child. This suggests that positive role models can have an effect on children, but that it is important to take into account aspects such as group membership.
The fourth study shows that households that gained access to clean water through Stockholm's expansion of filtered water pipes also had fewer infant deaths. The effect is entirely driven by infants and does not affect mortality among older children or adults. The effect is particularly large among families with middle and low socio-economic status, and among boys.
4. Contribution to the research frontier
Study one contributes to the research frontier by showing a relationship that has not previously been shown with representative and large-scale data. Since previous research tends to show that women work less after industrialization, this study can contribute new insights and show how big cities gave women an exception to the more general trend. Furthermore, it is one of the few studies that can estimate social intergenerational mobility for women in historical data. We take advantage of the unique opportunity in Swedish censuses to link women over time, so that they can be followed over the life cycle. Study two further contributes to the understanding of women's early labor force participation by studying not only the effect of large cities, but the importance of social networks within the city.
Study three contributes new knowledge regarding how interactions across class boundaries can affect children's life outcomes, by exploiting the fact that cities and individual buildings historically exhibited significantly lower levels of socioeconomic segregation. One difficulty in identifying how childhood environment affects children's life outcomes is that families sort into different neighborhoods and homes, which limits interactions across socioeconomic class boundaries. A few studies have therefore offered poor families the opportunity to move to "better" neighborhoods and studied how such a move affects children's outcomes. A limitation of this type of study is that it is difficult to show whether this type of intervention can be scaled up.
Study four contributes with identification and precision in the estimates that previously could not be achieved. The most cited studies on this topic have exclusively used aggregated data at the parish level and have thus had difficulties isolating the effect of water access from socio-economic variables and other health interventions. This study is therefore the first of its kind.
5. New research questions
In the course of the work, a number of new questions and research ideas have emerged.
A new subproject studies the diffusion of ideas in urban environments using Rotemans data. By linking the Rotemans archive and censuses with our own detailed patent data, we focus on technological innovation. The project is still in the development stage, but preliminary results show that inventors who moved to the city were more productive and collaborative compared to inventors outside of urban environments. By observing inventors before and after moving, we try to distinguish the part of this relationship that can be explained by selection. In a second part of the project, we intend to exploit variation in exposure to other inventors within the city and thus study spatial patterns and how neighboring inventors may influence each other. The project also aims to highlight the underrepresented group of female inventors (5% in our data), most of whom are located in Stockholm. It is of particular interest to explore possible networks of highly qualified women in the city and how these may have played a crucial role in encouraging female participation in innovation.
Another new sub-project studies social networks and their role in emigration. The project is based on the fact that chain migration is an important element in migration patterns over time and space. However, the existing research literature has been limited by having to rely mainly on either aggregated time series data or individual data with limited spatial resolution. Using our unique spatial data for Stockholm linked to migration registers that researchers have used in other studies, we can study the importance of social networks within a city with an unusually high level of detail. Advantageously, the Rotemans archive coincides with the extensive Swedish emigration to North America, which allows us to study migration patterns during a period of significant emigration.
6. Dissemination of research results
Our primary method of disseminating our research results is through publication in scientific journals. The first study has been submitted to the journal for review and the fourth study will be submitted shortly. The remaining studies will successively be sent for review when they have been completed. In addition, the project's research has been presented at scientific conferences and seminars in Uppsala, Stockholm, Copenhagen, Lisbon and Barcelona. The project also organized a workshop in economic history hosted by Uppsala University on 25-26 May 2023, with Professor Hans-Joachim Voth of the University of Zurich as an invited speaker.
Websites
Berger https://thorberger.wordpress.com/
Karadja https://sites.google.com/site/mounirkaradja/
Prawitz https://sites.google.com/site/erikprawitz/
Önnerfors https://www.su.se/english/profiles/maon7550-1.671839
During the Industrial Revolution, Stockholm evolved from a small capital city with 50,000 inhabitants to a modern metropolis with several hundred thousand inhabitants. The city grew largely through the migration of both women and men from the countryside. This project uses a newly digitized and geolocated database covering the entire population of Stockholm 1878-1926 (the Roteman database) combined with, among others, the 1880-1930 censuses to investigate the effect of the move to Stockholm on migrants and how residential and childhood neighborhoods within the city affected the future life outcomes of Stockholmers.
Background and summary of the project
1. Purpose
The purpose of the project was to investigate how the emergence of large cities during the industrial revolution affected individuals' labor market outcomes, and to study how growing up in cities affects the local population in terms of labor market, family formation, health, and political participation.
2. Development of the project
The project has resulted in four studies. The first documents that large cities were an early contributor to women's entry into the labor market. Already around 1910, women's labor force participation was much higher in the largest cities in the US, UK and Sweden compared to rural areas and smaller cities. The study analyzes the issue further by examining the labor force participation of female migrants to Stockholm compared to their sisters migrating to other locations. The study also examines the effect on marriage, fertility, health, and political participation. We also examine how migration to Stockholm during industrialization affects intergenerational mobility, i.e. the extent to which children's income correlates with their fathers.
The second study examines the effect of social networks on whether women arriving in Stockholm start working. The study compares women who arrive in the same year and in the same area of Stockholm, but find housing in different buildings. It then examines how the composition of neighbors affects the likelihood of working, with a particular focus on the number of working women and men in the building. We also examine whether individuals seek out specific occupations held by their neighbors.
The third study examines the effect of childhood environment in Stockholm using detailed data on housing and neighbors during the first 16 years of life. We exploit the fact that cities were historically segregated vertically (within buildings) rather than horizontally (between buildings) as today. This meant that many poor children grew up in buildings with highly educated neighbors. We study the extent to which such potential interactions affected the life outcomes of boys and girls.
The fourth study examines the impact of Stockholm's expansion of water pipes that brought clean water to connected buildings. The study examines how access to clean water affected child mortality. Through access to individual data, we can also distinguish effect sizes based on household characteristics such as socioeconomic status. The study is almost ready to be submitted to a journal.
3. Key findings
Study one shows that large cities were extremely important for women's early entry into the formal labor market. Unlike men, who had equally high employment rates regardless of the population size of their place of residence, it was mainly large cities that gave women this opportunity. We also find that working women in large cities refrained from marrying and having children, which was probably a requirement to be able to work to a greater extent. Furthermore, we find that Stockholm migrants exhibit very high social mobility, with the correlation between father's and child's income being significantly lower than among individuals who do not move. This shows that urban migration largely broke the link between parents' and children's life outcomes.
The second study finds that social networks were important contributors to women's labor force participation. We find that women whose first residence was in a building with more working women are also more likely to work, both during their first year in Stockholm and after 5 years in the city.
The third study shows that the neighborhood has an important impact on individuals' future working life. In this study, we find that children of parents with blue-collar occupations are positively affected by having more adult neighbors with high occupational complexity. In particular, we find that this effect is mainly driven by having neighbors of the same gender as the child. This suggests that positive role models can have an effect on children, but that it is important to take into account aspects such as group membership.
The fourth study shows that households that gained access to clean water through Stockholm's expansion of filtered water pipes also had fewer infant deaths. The effect is entirely driven by infants and does not affect mortality among older children or adults. The effect is particularly large among families with middle and low socio-economic status, and among boys.
4. Contribution to the research frontier
Study one contributes to the research frontier by showing a relationship that has not previously been shown with representative and large-scale data. Since previous research tends to show that women work less after industrialization, this study can contribute new insights and show how big cities gave women an exception to the more general trend. Furthermore, it is one of the few studies that can estimate social intergenerational mobility for women in historical data. We take advantage of the unique opportunity in Swedish censuses to link women over time, so that they can be followed over the life cycle. Study two further contributes to the understanding of women's early labor force participation by studying not only the effect of large cities, but the importance of social networks within the city.
Study three contributes new knowledge regarding how interactions across class boundaries can affect children's life outcomes, by exploiting the fact that cities and individual buildings historically exhibited significantly lower levels of socioeconomic segregation. One difficulty in identifying how childhood environment affects children's life outcomes is that families sort into different neighborhoods and homes, which limits interactions across socioeconomic class boundaries. A few studies have therefore offered poor families the opportunity to move to "better" neighborhoods and studied how such a move affects children's outcomes. A limitation of this type of study is that it is difficult to show whether this type of intervention can be scaled up.
Study four contributes with identification and precision in the estimates that previously could not be achieved. The most cited studies on this topic have exclusively used aggregated data at the parish level and have thus had difficulties isolating the effect of water access from socio-economic variables and other health interventions. This study is therefore the first of its kind.
5. New research questions
In the course of the work, a number of new questions and research ideas have emerged.
A new subproject studies the diffusion of ideas in urban environments using Rotemans data. By linking the Rotemans archive and censuses with our own detailed patent data, we focus on technological innovation. The project is still in the development stage, but preliminary results show that inventors who moved to the city were more productive and collaborative compared to inventors outside of urban environments. By observing inventors before and after moving, we try to distinguish the part of this relationship that can be explained by selection. In a second part of the project, we intend to exploit variation in exposure to other inventors within the city and thus study spatial patterns and how neighboring inventors may influence each other. The project also aims to highlight the underrepresented group of female inventors (5% in our data), most of whom are located in Stockholm. It is of particular interest to explore possible networks of highly qualified women in the city and how these may have played a crucial role in encouraging female participation in innovation.
Another new sub-project studies social networks and their role in emigration. The project is based on the fact that chain migration is an important element in migration patterns over time and space. However, the existing research literature has been limited by having to rely mainly on either aggregated time series data or individual data with limited spatial resolution. Using our unique spatial data for Stockholm linked to migration registers that researchers have used in other studies, we can study the importance of social networks within a city with an unusually high level of detail. Advantageously, the Rotemans archive coincides with the extensive Swedish emigration to North America, which allows us to study migration patterns during a period of significant emigration.
6. Dissemination of research results
Our primary method of disseminating our research results is through publication in scientific journals. The first study has been submitted to the journal for review and the fourth study will be submitted shortly. The remaining studies will successively be sent for review when they have been completed. In addition, the project's research has been presented at scientific conferences and seminars in Uppsala, Stockholm, Copenhagen, Lisbon and Barcelona. The project also organized a workshop in economic history hosted by Uppsala University on 25-26 May 2023, with Professor Hans-Joachim Voth of the University of Zurich as an invited speaker.
Websites
Berger https://thorberger.wordpress.com/
Karadja https://sites.google.com/site/mounirkaradja/
Prawitz https://sites.google.com/site/erikprawitz/
Önnerfors https://www.su.se/english/profiles/maon7550-1.671839