How can EA bring about the social change it desires? An analysis of historical social change portfolios

[James Herbert]

Executive Summary

This post aims to provide some initial insights into how EA might think about constructing its social change portfolio. Using a very dodgy methodology that relies far too much on ChatGPT, I examined a set of historical social change events and movements and identified their primary and secondary approaches to social change. A qualitative evaluation was carried out (using ChatGPT again) to gauge the success of these examples based on a pre-defined criterion. While the most frequently observed strategies were Social Movement Support (SMS) and Field Building (FB), the findings are preliminary and should be interpreted with extreme caution. They offer a starting point for EA to consider how to approach its social change initiatives. My hope is that this poor piece of research inspires someone to do it properly.

Introduction

This post offers a preliminary analysis of historical case studies of social change to inform EA community building strategy.

In a previous post, I described EA, broadly speaking, as an attempt at social change, and I suggested we need to think carefully about how we allocate our ‘social change portfolio’ between the following approaches:

• Social movement support

• Field building

• Network development

• Promoting the uptake of practices by organisations

In this post, I attempt to inform this thinking by looking at a set of historical case studies and analysing how their social change portfolios were allocated. I did this using ChatGPT.

Four approaches to social change

As discussed in my previous post, there are four primary approaches to social change:

1. Social Movement Support (SMS): Focuses on grassroots mobilisation and public campaigns.

2. Field Building (FB): Aims to create an academic or intellectual foundation for the movement.

3. Network Development (ND): Involves building relationships with other organisations, academics, and policymakers.

4. Promoting the Uptake of Practices by Organizations (PUPO): Focuses on influencing policies, regulations, and practices within existing organisations.

Historical Examples and Approaches Employed

Mark Lutter, founder and Chair of the Charter Cities Institute, put together this list of historical examples of social change.

I asked ChatGPT to identify the primary and secondary approaches employed by each movement or event.

This resulted in the following table (which I have not checked):

Historical Example	Primary Strategy	Secondary Strategy
Fabian Society	FB	ND
Corn Laws Repeal	SMS
YIMBYs	PUPO	SMS
Mont Pelerin Society	FB	ND
Meiji Restoration	FB	PUPO
Prohibition	SMS	PUPO
Progressivism	SMS	FB
Abolitionism	SMS
African decolonization	SMS	ND
Creation of Soviet Union	SMS	FB
Roman Republic to Roman Empire	FB
Bretton Woods	PUPO
Tiananmen Square	SMS
French Revolution	SMS	FB
Environmentalism	SMS	PUPO
Federalist Society	ND	PUPO

Tallying the results, you get the following:

• SMS: 10 primary, 2 secondary → 12 total

• FB: 6 primary, 2 secondary → 8 total

• ND: 3 primary, 2 secondary → 5 total

• PUPO: 4 primary, 4 secondary → 8 total

From the tallies, the “average” social change portfolio for the historical examples would look something like:

• Social Movement Support (SMS): 40% (12 out of 30)

• Field Building (FB): 27% (8 out of 30)

• Promoting the Uptake of Practices by Organizations (PUPO): 27% (8 out of 30)

• Network Development (ND): 17% (5 out of 30)

This suggests that the most frequent strategy among these historical examples is Social Movement Support, closely followed by Field Building and PUPO. Network Development, while significant, is less frequently observed.

The relationship between social change ‘success’ and the social change portfolios employed

I then asked ChatGPT to describe the relationship between social change success and the social change portfolios employed.

Defining ‘success’

I used the following definition of success: “A change of at least 15% on a scale of observed behaviour or beliefs that occurs in more than 25% of the population, and has consequences that persist for at least 10 years”

ChatGPT did a qualitative assessment and provided the following rundown. I have not checked this.

Fabian Society:

• Impacted intellectual and political elites more than 25% of the population.

• Long-lasting impact but not meeting the scale criterion.

Corn Laws Repeal:

• Significant shift towards free trade, likely meeting the 15% change.

• Impacted a majority of Britons.

• Lasting impact on trade policy.

YIMBYs:

• Still a recent movement; longevity is not confirmed.

• Scale and population impact are also not clear yet.

Mont Pelerin Society:

• Influential but primarily among intellectual and political elites.

Meiji Restoration:

• Massive societal and behavioural changes well beyond 15%.

• Impacted virtually the entire population of Japan.

• Consequences persisted for decades.

Prohibition:

• Huge immediate change in behaviour (alcohol consumption).

• Affected a vast percentage of the U.S. population.

• Repealed and thus did not persist for more than 10 years.

Progressivism:

• Significant changes in policy and public attitudes.

• Broad impact across American society.

• Many reforms persist to this day.

Abolitionism:

• Dramatic shift in beliefs and behaviours regarding slavery.

• Affected more than 25% of populations in countries where it was prominent.

• Lasting impact, slavery is universally considered a moral abomination today.

African Decolonization:

• Massive shift in governance and self-perception.

• Affected the entire populations of newly independent countries.

• Lasting impact.

Creation of the Soviet Union:

• Radical change in economic and political structures.

• Impacted the entire population.

• Consequences have persisted even after its collapse.

Roman Republic to Roman Empire:

• Significant changes in governance and social structure.

• Affected a large part of the Roman population.

• Lasting impact on the history of governance and law.

Bretton Woods:

• Influential but mostly among countries and their economies, not necessarily more than 25% of any population directly.

Tiananmen Square:

• Significant as a symbol but did not result in a 15% shift in behaviour or beliefs within China.

• Did not affect more than 25% of the population in the long term.

French Revolution:

• Drastic changes in governance and social structure.

• Affected almost the entire French population and beyond.

• Lasting impact on politics and philosophy.

Environmentalism:

• Significant shift in attitudes towards environment and conservation.

• Broad impact globally.

• Lasting impact and ongoing.

Federalist Society:

• Influential in legal circles but not necessarily affecting more than 25% of the U.S. population directly.

• Lasting impact on American judiciary.

The winners

Based on these qualitative assessments, ChatGPT suggested the following examples be coded as successful:

• Corn Laws Repeal

• Meiji Restoration

• Progressivism

• Abolitionism

• African Decolonization

• Creation of the Soviet Union

• Roman Republic to Roman Empire

• French Revolution

• Environmentalism

The popularity of the different approaches amongst successful examples of social change

I then asked ChatGPT to say something about the social change portfolios employed by this list of successful examples:

• Social Movement Support (SMS): Employed in 7 out of 9 cases, or approximately 78%.

• Field Building (FB): Employed in 5 out of 9 cases, or approximately 56%.

• Network Development (ND): Employed in 1 out of 9 cases, or approximately 11%.

• Promoting the Uptake of Practices by Organizations (PUPO): Employed in 2 out of 9 cases, or approximately 22%.

Discussion

The analysis presented in this paper offers a preliminary exploration into the allocation of social change portfolios based on historical case studies. The aim is to inform EA community building strategy by understanding the strategies employed by successful social change movements. In this discussion, I delve into the implications of the findings, their relevance to EA, and acknowledge the study’s limitations.

Strategy allocation in successful social change

The analysis suggests that Social Movement Support (SMS) and Field Building (FB) are the most frequently employed strategies in successful historical examples of social change. This finding implies that grassroots mobilisation, public campaigns, and the establishment of intellectual foundations are key drivers of successful social change efforts.

Field Building’s prominence among successful cases underscores the importance of establishing a rigorous intellectual foundation for a movement. This approach helps provide coherence, mobilise intellectuals, and leave a lasting impact on policies and societal norms.

Network development and promoting the uptake of practices by organisations

The analysis suggests that Network Development (ND) and Promoting the Uptake of Practices by Organizations (PUPO) were less frequently observed in successful social change initiatives. ND was employed in only 11% of successful cases, while PUPO was used in approximately 22% of them. This indicates that, while these strategies play a role in social change, they may be less central to achieving broad and lasting impact compared to SMS and FB.

The lower frequency of ND may be attributed to the historical context of many of these cases, where grassroots movements and ideological shifts played a more pivotal role. However, it’s important to acknowledge that ND can be crucial in building alliances, securing resources, and gaining political influence in contemporary social change efforts.

The moderate presence of PUPO among successful cases highlights the significance of influencing policies, regulations, and practices within existing organisations. In cases like Progressivism and Environmentalism, influencing the practices of organizations and institutions has led to significant, lasting changes in society.

The relationship between social change ‘success’ and the portfolios employed

The definition of success used in this analysis, involving a 15% change in behaviour or beliefs occurring in more than 25% of the population with consequences persisting for at least 10 years, provides a structured framework for evaluating the impact of social change initiatives. However, it is essential to recognise that this definition may not capture all aspects of success, especially in cases where long-term consequences are harder to measure or where shifts in behaviour and beliefs are gradual.

The qualitative assessments of success for historical examples add depth to the analysis but introduce subjectivity. Applying a uniform criterion to vastly different historical contexts can be challenging, and there may be cases where the interpretation of success varies. Nevertheless, these assessments provide an initial step towards evaluating the effectiveness of social change strategies.

Limitations

1. I am not a researcher and I have no idea what I’m doing.

2. ChatGPT is very unreliable and I haven’t checked anything.

3. This is a weird dataset.

4. This is a very basic methodology and it is probably fundamentally flawed.

5. Maybe ChatGPT misinterpreted my definition of success. I would argue neoliberalism counts as a success using my criterion, but ChatGPT seemed to disagree.

Conclusion

The analysis provides some preliminary indications that strategy selection may play a role in the efficacy of social change initiatives. While Social Movement Support (SMS) and Field Building (FB) appeared more frequently in what were qualitatively assessed as successful cases, it’s important to exercise caution in drawing firm conclusions. This is a very scrappy way of trying to answer this question, implemented by someone who doesn’t know what they’re doing. For EA, these findings could serve as an initial step in thinking more deeply about how to allocate resources across different strategies. For me, it is a point of evidence in favour of keeping the social movement part of EA strong. However, more rigorous research done by someone who actually knows what they’re doing is needed to substantiate these early observations and tailor them to the specific goals and challenges that EA aims to address. Please, someone do this for me.

[Bridget Loughhead]

Really interesting way to start looking at this idea. I have been wanting to think about this topic for a while but it’s such a large question — using ChatGPT seems like a good way to get a foothold.

Some other organisations that I think have lessons for EA are the 20th century “service groups” like Rotary International, Lions Club etc. Particularly Rotary has some similarities (local groups, international network) and some successes in ambitious projects (eradicating polio). Under the categories you put forward, I think this would be categorised under social movement support but perhaps also network development by influencing people who end up holding power.

It’s not clear to me what’s most useful about doing these comparisons — maybe we can see certain outcomes we want to avoid, or learn about ways that the initial philosophy of the organisation changes over the lifespan. I also think there are some interesting observations about what gives an organisation a long lifespan or do these movements primarily confine themselves to people of a single generation.

[Vaipan]

Interesting! Just to clarify for others, SMS means: A social movement is a loosely organized effort by a large group of people to achieve a particular goal, typically a social or political one.

Which isn’t what EA does, as it is much more field-building oriented. However SMS seems the most efficient way of obtaining positive results; do you suggest that we do the same/​try to think more broadly about public advocacy? Thinking about AI safety? Holy Elmore’s piece on AI safety advocacy supports your conclusion.

I also wonder how much this analysis is applicable for EA since EA is highly peculiar; it doesn’t obey the same rules that usually apply to communities IMO. So the ND seems especially relevant for EA since it’s basically working through a network of white, STEM, rich men (more than 60% of the community as per the last RP survey) who want to make a change. As long as EA doesn’t openly make efforts to become more diverse and more accessible, will it be efficient to go for SMS?

I might have misunderstood completely your post though!

[James Herbert]

Thanks for the comment!

I’d say EA is very much a movement.

First, MacAskill mentions it in his definition of EA:

(i) the use of evidence and careful reasoning to work out how to maximize the good with a given unit of resources, tentatively understanding ‘the good’ in impartial welfarist terms, and

(ii) the use of the findings from (i) to try to improve the world.

(i) refers to effective altruism as an intellectual project (or ‘research field’); (ii) refers to effective altruism as a practical project (or ‘social movement’).

Second, you can see it reflected in CEA’s programming. I’d say their Groups programme is a social movement support programme (and, as I’m funded by that programme, I’d say most of my work is social movement work).

Third, the outside world perceives it as a movement. See the Wikipedia definition, this New Yorker article, and this TIME article.