Campaign 08 Blog logo Campaign 08 BlogPolitics meets the social web
Political Ads

Political Text Messages and Who Pays to Send Them

Eleven days before the last general election my phone took forty one political text messages. Four were from campaigns in states I have never lived in. Two addressed me by a name that is not mine.

A phone on a desk showing a stack of campaign text messages beside a volunteer laptop and a printed voter list

Eleven days before the last general election my phone took forty one political text messages. Four were from campaigns in states I have never lived in. Two addressed me by a name that is not mine. One arrived at 06:12 on a Sunday, which I took personally. None of them told me anything I did not already know, and all of them cost somebody money.

That is the thing worth sitting with. Texting looks free because it is free to you and to me, and the simulators I spend my weekends in treat it that way: a cheap option on the voter contact menu, pennies a head, scaled up with a slider. The real version has a price list, a registration process, a carrier that can silently drop your whole send, and a legal distinction that decides whether a message is lawful or a five figure problem. Here is how the money actually moves.

Two completely different things both called texting

Everything in this area depends on one split that the headlines rarely make. A broadcast send, sometimes called A2P or application to person, pushes thousands of identical messages from a platform with no human pressing anything per message. Under the Telephone Consumer Protection Act in the United States, that style of send to a mobile number needs prior express written consent. No consent, no send, and the statutory damages run from 500 to 1,500 dollars per message.

Peer to peer texting is the workaround, and it is a real one rather than a loophole dressed up. A volunteer sits at a web interface, sees one voter at a time, and clicks send for each. Platforms such as Hustle, ThruText and Spoke are built around that click. Because a person initiates every message, campaigns argue the consent rule does not bite the same way. The cost of that argument is throughput: a trained volunteer gets through roughly 300 to 600 first messages an hour, and then has to answer the replies, which is where most of the actual labour sits.

So the choice is not cheap versus expensive. It is legally exposed and instant versus legally defensible and slow, and campaigns pick based on their lawyer, not their budget.

What a send actually costs

Numbers from vendor rate cards and public filings, which move around but not by much:

  1. Peer to peer platform fees run about 1 to 4 cents per outbound segment, often with a monthly platform minimum of 500 to 2,500 dollars. A message over 160 characters bills as two segments, which is why so many political texts end abruptly.
  2. Picture messages cost roughly 2 to 4 times a plain text, so a campaign that attaches a candidate photo has quadrupled the line item for a thing most recipients never load.
  3. Carrier pass through charges on 10DLC registered traffic add a fraction of a cent per message, and a one off registration fee of 4 to 50 dollars per campaign brand.
  4. Toll free or short code routes cost more up front, with short code leasing around 500 to 1,000 dollars a month, and are mostly used by national committees that send continuously.
  5. The invisible cost is the phone number data. Matched cell numbers from a commercial voter file vendor run something like 2 to 10 cents per record, and match rates of 60 to 70 percent on a registration list are considered good.

Put together, a 200,000 message programme is not the 2,000 dollars the per message rate suggests. With data, platform minimum, picture upgrades and the staff running the reply queue, 12,000 to 20,000 dollars is closer, and the fundraising texts that fund it have their own processing cut on top.

The carrier is the referee nobody mentions

Here is the mechanic that would make a decent game system and that no simulator has touched. Between the campaign and the voter sits a mobile carrier running spam filters, and the carrier does not have to tell you when it drops your traffic. A send can report as delivered at the platform layer while a large share never lands on a handset.

The registration regime called 10DLC exists partly to manage this. Campaigns register a brand and a use case against their sending numbers, get assigned a trust score, and that score shapes how much volume per minute the carriers will accept. An unregistered number sending political content at volume in late October gets throttled or blocked within hours. Links are the other trigger. A shortened link on a shared domain is a classic spam signature, which is why serious operations buy their own branded short domain and warm it up for weeks.

I have watched a volunteer team work an evening shift convinced the list was cold, because reply rates had collapsed. The list was fine. One of their three sending numbers had been flagged, and nobody looked at the per number delivery breakdown until the next morning.

Why the cheap and effective story does not hold up

The usual claim, repeated in consultant decks since about 2018, is that texting is the highest return voter contact method available because it costs a fraction of a mail piece and gets read almost immediately. The read rate part is true. The rest deserves a harder look.

Field experiments on text based get out the vote efforts, including work published by the Analyst Institute and several academic replications, generally find turnout effects in the range of nothing to about 1 percentage point, with the larger effects concentrated in low information local races and among people who had some prior relationship with the sender. Cold texts to a purchased list sit at the bottom of that range. Compare that with in person canvassing, which the Gerber and Green work has repeatedly put around 3 to 8 percentage points per contact, and the efficiency argument gets narrower.

There is also a cost that does not show up on the invoice. Every unwanted text trains a voter to treat the channel as junk, and that damage is shared across every campaign that uses it. Opt out rates of 2 to 5 percent per send are normal, and those people are gone permanently, including from the fundraising list that was paying for the programme. A channel that degrades with use is a different asset from a channel that scales, and the sliders treat them as identical.

Honest limit on what I can tell you: the published experiments skew heavily toward American turnout work funded by one side of the aisle, and the private vendor numbers nobody shares could point somewhere else. I would not bet much on any single effect size here.

How to work out who texted you

Four steps, roughly five minutes, and they work more often than you would expect. Start with the disclaimer in the message itself, usually a committee name crammed into the last twenty characters, then run that name through the relevant filing database. Our walkthrough on how to read a political ad disclaimer covers the committee types you will hit.

Second, look at the sending number. A ten digit local number points to a peer to peer programme, a five or six digit short code points to a national operation with a leased code. Third, check the link domain against the committee website, because a mismatch often means a vendor or an unaffiliated group is behind it. Fourth, search the committee's disbursements for the texting vendor name. Spend on text platforms shows up as a line item, and the same technique used for tracking campaign spending in platform ad libraries applies to filings.

Replying STOP does work, and it works at the platform level rather than across the industry. You are off that committee's list, not off the sixteen others that bought the same data file. If you want the broader map of how these channels connect, the rest of the political ads coverage goes channel by channel.

Frequently asked questions

Broadcast sends to mobile numbers need prior express written consent in the United States, and sending without it carries statutory damages of 500 dollars per message, trebled for willful violations. Peer to peer sends, where a human clicks send on each message, are the route campaigns use to avoid that requirement, and the legal ground there has shifted with each court ruling since the Facebook v. Duguid decision in 2021.

How did campaigns get my mobile number?

Usually from a commercial voter file vendor that appended phone data to your registration record, which is public in most states. Donations, petition signatures and online form fills are the other common sources, and a phone number given to one committee is frequently shared across affiliated committees and sold onward.

Why do the texts get my name wrong?

Because the match between a voter record and a mobile number is probabilistic. A number reassigned to a new owner, a household where one line is shared, or a stale commercial append all produce confident messages addressed to the wrong person. Match error rates in the 10 to 20 percent range are not unusual on cold files.

Do any election simulators model texting properly?

Not that I have found. Democracy 4 and President Infinity both treat digital contact as a budget allocation with a smooth return curve. Nothing I have played includes consent requirements, carrier filtering, opt out attrition, or the volunteer hours that peer to peer sending actually consumes, and those four things are most of what makes the channel hard.

MA
Martin Ashby

Martin has played political strategy and simulation games for twenty years, from Diplomacy by mail to modern election simulators. He writes about what those games get right about campaigns and what they flatten.

More posts by Martin

More in Political Ads

01doorganico 084goad 3hot8 4465456 after5internprogram ahandfulofblessings ahmadpc ahomcybersecurity allodialresources annodex antiquesandperiodcollectibles aotc asianworldkashmir beedlectrics bloggertemplatefree bluefeatheragency bodyweightsolution carfixadvisor catstreetcollective cbstvsports cnfanspurchasing committeeof500yearsofdignityandresistance covertopsdjs cs-kaiyuangame cuestacomunicaciontotal cxoutlet dancebetgame deskgearhub destinyvillastay detectmagazine di33 diamondsteelbuildings donmo dyanadropship emarsysfinance emimattl escapechan eskitarih evautoplay everythingweddings-orlando farmzone feedviewerapp foxstudiosscratch gamevui123 gleasongrade greetingsday hacklabasu haskell iw-kaiyun jagoanproperti jilidskripsi joltbusinessfinance juegodepokergratis juegospucca kaeru-festival latestmehndidesigns lestanzediottaviano litemagazines localcommunityhero lustrsazan mariashebets menswheels newwebstechs1 nikkit ootheday oxigenvpn plurfitness prefetial probudgetcarpetcleaning psychicmrsmartin pyaten qvantor rejoignezlogis rexzbot riselayouts ruchin rwxqsgd5 sakaltime shibiaowang shop-bot skrtips spotlesshaven strcoding tetapatrium topselller uneplanteartificielle vipertechcr webtech1223411 wo-kygaming yamaha-mekar yeknava yoybuysheets zenunwa yamaha-mekar jilidskripsi ankertoto ankertoto ankertoto ankertoto