Letter writing (including email) and phone call campaigns are popular actions people are called to participate in to create pressure and change legislator behavior. It's taken as axiomatic that this is effective.

I've done my damnedest to search out empirical evidence that it is effective, however, but what few papers I can find on the subject directly use "political efficacy" as their measure. The problem being that political efficacy is someone's belief that their actions impact the behavior of their representatives, rather than measuring the impact itself.

This is problematic because I have found studies that measure the extent to which legislators confronted with constituents who hold different positions than their own tend to assume those constituents are ill-informed (to be polite), and that they tailor their communications to suit the constituent audiences with an eye towards persuasion/placation, and furthermore that this is effective. This makes constituent feelings of empowerment a rather dubious measure at best.

I assume it's a problem of using the wrong search terms, but this one's cooking my noodle. What studies have been done to try and actually measure the impact of constituent communications directly? What terminology do they use to describe it? Where are they published?

The answer I will ultimately click 'accept' on needs to show me academic-grade sources, but for now I'll take whatever people've got.

  • 15
    I'm not certain you're going to find this kind of data, on methodological problems. To get at the question you're asking, you'd have to (a) ask political representatives themselves, or (b) study the administrative flow within political offices to see if and how such messages are read, compiled, summarized, and ultimately presented to the political rep. The former is problematic, since reps have a vested interest in appearing to be attentive, regardless of the facts; the latter could be operationalized well, but is unlikely to be agreed to by reps (for privacy reasons). Commented Apr 14, 2021 at 17:53
  • 1
    You may end up being right, @TedWrigley. But I've been surprised at what studies have existed on other subjects, such as the measurable impact of voting in main elections on subsequent policy outcomes or how state legislatures' behaviors differ based on whether or not polling data is shared with them. The quality of these natural experiments may be debatable but in the absence of perfect information, I'm interested in seeing what the landscape of available information is - hence 'evidence' instead of 'proof.' Eventually I'll just explore the literature myself, but I need to find it first. Commented Apr 14, 2021 at 18:08
  • 1
    Also, is this US specific, or are you looking for examples worldwide?
    – divibisan
    Commented Apr 14, 2021 at 19:54
  • 1
    @divibisan Worldwide is fine, but it will need to be a representative democracy. Commented Apr 14, 2021 at 20:37
  • 1
    @dandavis There's a variety of ways to get at this information without asking the legislators, though it's all got validity issues of course. Consider Butler & Nickerson (2011 - DOI 10.1561/100.00011019) where they did a partially controlled experiment on exposing legislators to polling data. Commented Apr 17, 2021 at 2:21

2 Answers 2


It can be effective, but often is not. An occasion where it was, that happened to me.

Back in the early 2000s, I found an interesting little book on Amazon.co.uk, a study by the RAND Corporation of the UK's capabilities and infrastructure for building nuclear-powered submarines. It made a convincing case that if the sole shipyard capable of building them had long gaps between submarines, many of the skills involved were in danger of being lost.

A few months later, the UK Government announced that it was starting the process of designing replacements for the Vanguard-class submarines. That project is still underway, for what is now known as the Dreadnaught-class submarines. The Liberal Democrats opposed the project, suggesting that the existing submarines could have their lives extended at lower cost. That may well have been true, but it would increase the risk of losing the capability to build more submarines later.

It's a real political question if the UK should have nuclear submarines. But it would be really poor governance to accidentally lose the capability to build them. Giving up a capability like that should be done with a full understanding of its consequences. Since my MP at the time was a Liberal Democrat, I e-mailed him with that argument, and offered to loan him the book. He accepted with alacrity, and a couple of weeks later, the party stopped arguing for the life-extension programme.

Writing to representatives with solid factual information that they weren't aware of can be very effective. Saying you disagree with them is usually less effective.

  • 21
    Technically we only know correlation, not causation here, but that's still a pretty cool story (which I'm assuming is true). Commented Apr 24, 2021 at 20:47
  • So my personal idea I had years ago, of mailing a copy of a book on how to "easily" solve the USA's health care system to my senator might have an impact! Commented Jul 27, 2022 at 20:23
  • 1
    @MarkStewart Only if they are opposing healthcare due to a sincere belief that it can't be implemented. I suspect that in reality, they oppose it because they don't want citizens to have healthcare. Sending them information on how to provide citizens with healthcare won't help with that. Commented Apr 5, 2023 at 23:07

So I don't have empirical evidence that is systematic. But I do work for such a person.

In our organization, at least, we take the correspondence that comes in (email is best, and we strive to respond to each one we get, even if not individually or personally; we get thousands a month, and we have five people for whom it is part of their job to directly deal with communication from constituents) as a rough measure of public sentiment on a particular issue. We are aware that it is not a scientific sample.

But it is something of a measure of issue engagement. If you are motivated enough to write in about something, you probably care a lot about that (because most people do not write in). Engaged people, we assume, also tend to be voters and perhaps influencers of those in their network, so their opinion matters in that it may gauge the effect on re-election, for example.

I think you would be hard-pressed to get systematic empirical evidence on this question, although perhaps asking your friendly neighborhood political scientist would help. Best chance is that someone has been able to run a survey of correspondence managers.

  • could you email me at democracyusersmanual-at-gmail-dot-com if you'd be willing to take a brief interview? I suspect you're right, and in addition to being my own local political scientist, I've also nagged my colleagues to death. Which means I need to Scholar the F Up and do it myself, I guess. Commented Apr 27, 2021 at 13:32
  • 2
    I suspected you might be a political scientist; I am too. (Tribe identity markers and all.) I'd be happy to mail, but maybe not until the weekend coming? Long week of working for the people here... ;-)
    – Nate
    Commented Apr 27, 2021 at 13:57

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .