In recent years, we have witnessed several shifts in social norms and publicly expressed attitudes. In 2018, public opinion on Black Lives Matter changed from mostly skeptical, to mostly supportive. This tendency continued until May 2020, when the murder of George Floyd massively boosted the movement’s popularity, driving changes in attitudes towards the police, and in perceptions of the gravity of the problem of racial inequality in the US. Another social movement, #metoo, took off in 2017 following allegations against Harvey Weinstein, and brought about substantive changes in norms surrounding sexual harassment and sexual consent. And while both BLM and #metoo are intricately connected to the spread of social media, similar mass changes in attitudes took place also before the proliferation of Twitter. The figure below plots a measure of Americans’ attitudes towards same-sex relations. It demonstrates that public opinion became significantly more tolerant of homosexuality around 1990.
Every now and then, societies experience changes in patterns of behavior, attitudes and values. Some of these changes unravel slowly. Others, like the cases above, take place within the span of a few years, or even months. Some of them are transitory and quickly fade out; others signal the transition to new moral paradigms.
In the late 18th century, trade in sugar produced by slaves in Britain’s West Indian colonies accounted for a significant share of the empire’s commercial wealth. In the span of 46 years, a powerful movement for abolition managed to ban the slave trade and dismantle the institution of slavery in the colonies. Initially promoted by Quakers and other religious groups, abolitionism gradually garnered overwhelming popular support, with thousands signing petitions to ban the slave trade, and participating in boycotts of sugar products. Below is a graph of the cumulative number of petitions in favor of abolition over time (from an ongoing project with Valentin Figueroa I hope to talk about in a future post).
The abolitionist movement is an instance of change in moral values, with important institutional ramifications. Other instances of social change are reflected in cultural practices. Footbinding, the painful practice of tightly wrapping women’s feet in childhood to give them a lotus shape by adulthood, had become prevalent in China since the 11th century. Bound feet were not only considered aesthetically pleasing and indicative of high status, but also constituted essential prerequisites for a good marriage. Sometime in the 1870s the practice started losing in popularity. A custom that had persisted for nearly a millennium faded away entirely in less than a hundred years.
Another example: Between 1870 and 1900, France was the first of several European countries to experience a significant drop in the average number of children born per woman. This development, known as the fertility transition, was relatively rapid, has not been satisfactorily explained, but is thought to be of major historical significance – like the European Marriage Pattern discussed in Tracy’s latest post, it has been connected to the divergent developmental path of Northwestern Europe from the rest of the world.
Such discrete identifiable jumps in cultural behaviors, moral values and social norms raise two immediate questions: Why does social change happen? And when? What follows below is not the answer to these questions, but a few thoughts, and a loosely pieced together (almost certainly incomplete) set of insights from research in various disciplines that have touched upon the topic.
Fundamentals and superstructure
Many instances of value shift can be linked to changes in what we would term “fundamentals,” meaning tangible changes in a society’s material conditions. The movement for the abolition of slavery in England took off in the midst of the Industrial Revolution and the massive reorganization that industrialization brought about in the structure of labor relations and the relative power of different classes in society. Historians have generally identified a relationship between support for abolition and the rise of capitalism, even though they don’t always agree in the precise nature of the connection. In Capitalism and Slavery, Eric Williams pointed to the declining profitability of the slave trade as the relevant change in fundamentals – a thesis later rejected by others. For David Brion Davis the fundamental shift was more connected to the rise in power of a new industrial class that simply viewed the world in a different way compared to old elites. For industrialists, the legitimization of a nascent system of production reliant on free labor went hand in hand with the illegitimacy of the institution of slavery. Their rising economic power led to the gradual spread of their values to the rest of the population – what Davis calls “ideological hegemony.”
In the case of footbinding, one can identify a temporal correlation between the demise of the practice and China’s power decline and submission to Western imperialism after its defeat in the Opium Wars. Changes in various social groups’ relative status likely played a role in this case as well. The first to adopt unbound feet were Western-oriented educated Chinese elites, who resorted to commitment and coordination devices like anti-footbinding societies to help them abandon the old norm (as discussed by Gerry Mackie in his 1996 ASR article).
Change in the relative power of different groups – sometimes driven by changes in material conditions – is a pattern encountered in other instances of social change (see for instance the role of women’s labor during WWII for female empowerment and long-run changes in gender attitudes). In other cases, the change in fundamentals comes in the form of a new technology – Fernandez-Villaverde, Greenwood and Guner show that the introduction of the pill in the 1960s had a dramatic impact on attitudes towards extramarital sex.
Does that mean that, for every instance of social change, we should immediately look for the root cause in a contemporaneous shift in fundamentals? Two observations would complicate such a conclusion: first, the link between the material change and the associated cultural one is not always so obvious. Some social changes coincide with periods of economic change only if we define periods very broadly (i.e. in the same century) and the web of mechanisms connecting the two types of change remains opaque. Abolitionist sentiment grew during the Industrial Revolution, but what exactly was the channel connecting industrialization to attitudes towards slavery? Which economic shift drove the fertility transition (if interested, look here for an overview of competing economic explanations)? Second, there are instances in which massive shifts happen following what, in the bigger picture of things, is a small shock. Police brutality and unequal treatment of minorities was prevalent for a long time before the murder of George Floyd – but it was that isolated event that catalyzed a cascade of protests.
One characteristic of social change is that it is, well, “social” in nature. That means that interactions between individuals play a crucial role in determining whether the change will happen or not, as well as the speed with which it will happen. This is what is responsible both for the fact that social change is often only noisily connected to underlying changes in material conditions, and for the fact that social change can move at variable speeds, sometimes slowly and sometimes as a cascade.
For instance, assume people have a taste for conformity, but that this taste is not uniform across the population. When a new idea or behavior is introduced, it spreads from the least to the most conformists, slowly at first, but then increasingly rapidly (here is a model of such a process by Jim Fearon). In this classic paper by Bikhchandani, Hirschleifer and Welch, a similar effect can obtain, not necessarily because people are inherently conformists (which is not to say that fact cannot also be true), but because they want to do the right thing in any given situation, and others’ behavior serves as a signal as to what the right thing is. These are only two examples of a social propagation mechanism. What any such mechanism implies is that the probability anyone will adopt a new idea, value, or behavior depends on the share of others who have adopted it. This type of dynamic can produce two of the typical empirical patterns of social change: an S-shape (apparent in both of the graphs above), with prolonged periods of no or slow change followed by rapid acceleration; and tipping points, whereby, when a critical mass of adopters of the new norm has been achieved, everyone else adopts the norm as well.
Finally, there are social networks. There is no better demonstration of the presence of social multipliers, than the fact that network connectivity crucially determines the diffusion of social change. Recent historical work on this is exciting. Aidt, Leon and Satchell show that connectivity across locations drove the diffusion of the English swing riots of 1830-1831, over and above any similarities of locations in terms of fundamental socioeconomic factors. Garcia-Jimeno, Iglesias and Yildirim use exogenous variation in rail connectivity and in the telegraph network to identify how female-organized temperance protests spread in the US between 1873 and 1874. Becker et al. (2020) show how personal ties to Luther helped spread the ideas of the Reformation. These studies do not necessarily give us the whole picture of why these historical events happened in the first place, but they do tell a big part of the story of when, where and how of each of them unfolded.
Putting it all together
One takeaway of the ruminations above is that deep structural changes and social propagation mechanisms join forces to drive social change, in ways that are only partly predictable. Shifts in fundamentals may only manifest in behavioral patterns with a lag, because of social forces like conformism or preference falsification. Social propagation mechanisms can also drive social change when the underlying structural changes are not very large – with social forces at play, small shocks in fundamentals can have a big impact.
An exciting development for research in the crossroads of history and quantitative social science is the increasing availability of data and methods to trace the development of values and norms in the long-run. Some of the papers referenced above illustrate the promises of such approaches. Text digitization and analysis through machine-learning techniques now allows us to map the change in ideas and culture and link that to existing information on changes in fundamentals (see an example on mapping the evolution of gender and racial stereotypes). Such tools, in combination with the more traditional toolkits of history and formal theories of human behavior, open exciting new prospects for understanding social change, both past and present.