Are we underutilizing grassroots-style political advocacy?

TL;DR

EA might be unterutilizing direct contact of politicians through constituents, especially by phone calls, letters, and email. There is evidence that relatively little effort can have a significant effect, as long as voters call with a specific agenda, frame their concerns appropriately, and the topic is not highly polarized.

The Hypothesis

In this post I’m focusing on grassroots-style advocacy, where a number of people contact their respective representatives in order to nudge them into voting in a certain way with respect to specific legislations. I focus on calling, and writing emails and letters, or maybe even visiting politicians in their hometown offices; it’s not meant to be exhaustive, surely there’s more to investigate here.

I see three main settings: 1) parliament will vote on a law and you want to influence the politicians’ voting behavior; 2) parliament is planning on voting on a law, but the proposed law could use some improvements; and 3) you have an idea for a new law or a modification of another recent law. I think all three would need different considerations (and there are probably more), however, I will not differentiate between them—this is only meant as a short note.

I do realize that certain actors in the EA (and adjacent) space might actually use grassroots-style direct contact of politicians (especially in the animal advocacy space). However, it does not seem to get any attention in the “mainstream” EA community. A lot of people are interested in EA and want to participate in a meaningful way beside donating, but without making a complete career switch. I have the impression that the community is turning these people down, with the argument basically being that there’s not really that many effective things to do for volunteers. I think that giving this an “official stamp of approval” would make a huge difference (obviously conditioned on the EA community actually thinking that it is effective).

There has been some talk about what a “Task Y” could be, a task that is effective and helpful, where a lot of people can participate without too much time commitment, and that also gives them the feeling that they are really contributing something valuable. I have the impression that influencing your politicians might be a partial answer to this question.

Epistemic status

These are preliminary thoughts and I am looking for feedback. I spent a weekend going through some literature and sorting my ideas, but I don’t have any domain expertise in the area. I might be overly optimistic since I’m easily excitable.

Would this have any effect at all?

There is considerable evidence that politicians are influenced by direct contact of their constituents. Staffers and politicians are eager to claim that calls, letters, and emails do have an effect on their policymaking, and there exist many anecdotes of laws that have been passed and/​or modified based on constituent-input. There are some observational and even a few randomized studies indicating that these effects actually exist. In one special case, a small number of (randomized) calls to representatives led to a ~11-12 percentage point increase in voting in favour of a certain law compared to the control group. I’m far from calling these data conclusive yet, though.

Importantly, this all depends a lot on the kinds of laws under question. For highly partisan or polarized topics—immigration, guns, abortion, … - calling your representative will probably have no effect at all (exceptions exist, of course, but they usually require extreme amounts of constituents and media attention). Most evidence for effectiveness I found was focused on more innocuous topics, such as a school bullying law in response to a student suicide, or making certain government documents more user-friendly.

In essence: there is some evidence suggesting that contacting politicians on non-polarized topics should be reasonably effective.

Do these kinds of issues actually exist?

In topics such as health care reform and mental health, or in animal advocacy, there should be quite a number of laws that would have a considerable net-positive effect.

Two examples:

In early 2020, the German government voted on whether to make organ donation the default with the possibility to opt-out, vs making it opt-in, but asking citizens at some points when they visit a citizens’ office. I have the impression that the first option would be very much more aligned with EA-values—there’s practically no cost, everyone can still decide for themselves, and lots of lives would be expected to be saved (correct me if I’m wrong, please!). Although votes in the German parliament are usually divided along party lines, this one had members of all parties vote in all directions. The vote was relatively close, and the opt-in solution was accepted. A certain nudging of politicians might have helped change this outcome.

Another example might be to ban chick culling, so that it will be replaced e.g. by in-ovo sexing, a technology that as far as I can tell exists, but is not yet cost-effective enough. Most people (including meat eaters) don’t like the idea of killing billions of one-day-old chicks, it seems quite obviously EA-aligned, and a law banning (or heavily fining) the practice would probably speed up the wide-spread introduction of alternative technologies by many years.

On the other hand, I’m much more sceptical when it comes to long-termist causes, or highly complex topics where it really depends a lot on the fine print.

Is it really cost-effective?

I really don’t know yet. Some of the evidence suggests that even a few dozen calls can influence a single politician to a sufficient degree on certain topics, while others might need thousands or tens of thousands of calls and still might not have the desired effect. I would expect an average call, email or letter to cost anything between 5 minutes and two hours. A direct visit of a politician (e.g. at their local offices) might be a couple of hours. The effect if successful obviously depends on the law in question.

As a side note, I would also argue that 1,000 people spending one hour might have a very different value than 1 person spending 1,000 hours, i.e.: the cost of a project is not simply a linear function of the number of hours (and money invested etc), but depends on how these hours are distributed among different people. If we could mobilize 1,000 people to spend one hour, counterfactually they might have spent this one hour with free-time, they might gain a significant degree of meaning and community-involvement from this one hour and potentially making it quite positive for them; this increased involvement in the EA community might even lead to people identifying more with the ideals and donating more. The one person spending 1,000 hours to start a new project might have a very different experience, especially if the project turns out to be unsuccessful (which most projects probably are?).

One potentially substantial problem might be that many offices get a lot of calls, letters, and emails—numbers vary a lot, depending on source, political season, external events, nation & state, and many more, but one anchoring number might be several thousand contacts per “normal” week. Standing out among those many contacts might require a large amount of people, but I would expect this to depend also on the concrete issue.

All this does not yet include overhead costs—we would need to find useful laws to influence, we would need to organize the campaign, we would need to evaluate the effectiveness. We’d also need to incorporate potential negative effects, of course.

The right format

The main argument I’m making is with respect to calling, emailing or writing letters, or meeting politicians in person. The first three all have different pros and cons, with different sources citing different ones as the most effective. E.g. written letters and calls are more personal and hence might be more persuasive, but emails can be better indexed by the data processing software the politicians’ offices use (they seem to be very out-of-date).

Most sources discount online petitions as virtually worthless (although there might be exceptions if politicians can verify that it’s their very own voters who signed the petition). I found little up-to-date work on social media (esp. Twitter & FB), but the evidence that I found seemed quite negative. I’m not sure how demonstrations compare to these methods.

On small scale problems, organizing more along these lines might also be feasible.

Different countries, different rules

Effectiveness and cost-effectiveness might be quite different for different democracies. E.g. the US system makes politicians quite responsible to their constituents, so they might be very inclined to listen to them. In Germany by contrast, decisions are usually made on the level of parties; however, I would assume that fewer people might be actually contacting their representatives, making it easier to cut through the noise.

What are the potential risks?

Two general policy problems are that a) we might inadvertently be supporting the wrong side of an argument (or politicians might misunderstand what we are saying); and b) we might divert attention from more cost-effective and more relevant, but a-little-harder-to-achieve causes/​methods by luring EAs into a less-than-optimal easy-win area. I do think this would require some more detailed and quantitative analysis of the evidence, and potentially some pilot tests. In general, many of the risks mentioned here apply to some degree, but I also think that some of them might be less pronounced since we’re talking about smaller forms of more indirect nudges.

Most other risks I see are related to the (cost-)effectiveness. For one, this approach might require too many people to be involved, and too much time commitment of those (they would need to regularly contact their officials). This also expands to the potentially large amount of organizational overhead. This effect is exacerbated by the fact that different countries have very different political mechanisms—each country would require independent work.

On the other hand, we might not find sufficiently many laws/​low-hanging fruits worth promoting—maybe they just don’t exist (policy is hard), or they are too hard to find.

I also imagine the whole process to be relatively hard to quantify; although I can think of several techniques on how to gather semi-reliable data, it’s by no means trivial and would require non-negligible amounts of empirical research. This also entails the possible failure to realize so if it turns out to not be an effective way of improving policies.




I have many more thoughts, but this post is already way longer than I had planned for it. I’m interested in what others think about the whole topic, whether I’m making any inferential mistakes, whether I’m overlooking any opposing evidence, or if people would be interested in exploring this further.