I recently explained that you can help increase voter turnout this year simply by throwing a friendraiser. There were some questions. Here are some answers..

Question: What is a friendraiser?

Answer: A friendraiser is a chance for you to use those party throwing skills you cultivated in college to (a) throw a great party, (b) see all of your friends, and (c ) raise some money for a great cause that you care about. If you’re nervous about asking your friends for money, don’t be! They see you asking them to support something that’s important to you. That’s what matters most.

Throwing a friendraiser is easy! We’ll walk you through it.

First, decide what type of party to throw.

As far as we are concerned, this is the hardest step. Should you throw a raging keg party? A quiet dinner party? A sassy cross stitch party? Really, there are so many great types of parties. Pick the most fun idea and go with that.

Second, pick a date.

As soon as possible is the best time to throw a party. This weekend would be great, actually. The impact of donated dollars diminishes as we get closer to Election Day. If at all possible, try to throw your party before National Voter Registration Day (September 27, 2016).

Here are some election-related dates for you.

• Monday, Sept. 26, 2016 — First presidential debate
• Tuesday, September 27, 2016 — National Voter Registration Day
• Tuesday, Oct. 4, 2016 — Vice presidential debate
• Sunday, Oct. 9, 2016 — Second presidential debate
• Wednesday, Oct. 19, 2016 — Third presidential debate

Third, invite all of your friends.

Set up a Facebook event, use Google calendar, send emails/texts/carrier pigeons, etc. Really, whatever floats your boat. Just remember to be upfront about the fact that you’ll be asking them to support Vote.org at the party.

(In case you’re curious, the the VoteAmerica team always uses Google meeting requests to schedule parties, and we are a notoriously fun group of people.)

Fourth, set aside time at the party to make a short pitch.

Don’t worry, we have some words for you (below), but you should make it as personal as you can. People love that shit. Do you care about what the outcome of the election is? Why? Why do you think more voting will make a difference? What was it like the very first time you ever voted? Really, people are there to support VoteAmerica because you support VoteAmerica. Tell them why.

Fifth, have people verbally commit to donating. (Optional but epic step)

Social validation is a really great motivator when it comes to donating. So what you want to do is line up a couple of friends beforehand who you know aren’t shy about going first. These people are plants (in a good way) Then after everyone has stated why voting is important to them, you start asking people to commit to donating. Your plants pipe up and immediately pledges a reasonable amount, like $25 dollars. This will get things going!

• Sidebar: It is very important that the plant starts with a reasonable donation. I was at a fundraiser recently where the plant opened with a $1000 donation. This does not create momentum and excitement. This makes people feel like their donations are not generous enough. We want to start small and build from there!

Here’s how this would work at my party:

• Me: “I’m glad you could all make it to my friendraiser. VoteAmerica is on track to run one of the biggest Get Out The Vote efforts in the country. Let’s help them by raising some money. Who’s with me?”
• Toby the cat: “I’ll chip in ten dollars!”
• Mona the cat: “I’ll chip in ten as well!”
• Human friend: “Well if the cats are doing ten, I’ll do twenty!”

Sixth, guide everyone to the nearest laptop(s).

We use ActBlue for fundraising. Here’s the URL for our fundraising page:

• https://secure.actblue.com/donate/voteamerica

If you want to know how many people donated because of you and your AWESOME PARTY you can add a referal code to the URL. Here’s how that works:

• https://secure.actblue.com/donate/voteamerica/?refcode=friendraiser-YOUR-NAME-HERE

Important things to note:

• VoteAmerica is a 501(c)(3) nonprofit charity. All donations are fully tax-deductible.
• Actblue sends automatic “thank you” emails and receipts, so people are good come tax-time

Also, put out a jar and drop a $20 bill in it before everyone gets there. Some people still use cash. Crazy, I know. Anyway, print out VoteAmerica logo and tape it on the jar. Or get creative:

Last! But important!

• Post friend-raiser, send a thank you email to everyone for attending.
• Make sure to include a link to the donate link that you made (above) in the email. This way anyone who forgot to donate at the party (or who is inspired to donate again, it happens!) can still kick in some dollars.

Now sit back, relax, and know that you’ve done more than 96% of people to increase voter turnout.

Tips on how to make the pitch at your party

• First, ply everyone with alcohol.

• Second, tell everyone what inspired you to host a friend-raiser for VoteAmerica. Do you care about the outcome of the Presidential election? Why? Why do you think more voting will make a difference? What was it like the very first time you ever voted? Really, people are there to support VoteAmerica because you support VoteAmerica, so tell them why. Here’s what a long-time VoteAmerica supporter had to say recently: “I will never forget my first time voting. I was a nervous 18-yr old, accompanied by my mom, and I voted for Bill Clinton. My heart was going to burst out of my chest when I submitted my ballot. It was a big day. I’ve been watching the attacks on voter access the past few years, and part of the reason I support VoteAmerica is that it makes voting easier, not harder. That’s the way it should be. Every voter should be able to feel that heart-bursting feeling.”
• Third, tell attendees some basic VoteAmerica facts and explain how you became involved. You might want to mention that VoteAmerica ran spent over $1 million dollars on the largest ever HBCU mobilization drive in history in 2020. Or that VoteAmerica sued Georgia and Kansas in 2021 and 2022 and won both lawsuits. Or that the VoteAmerica team keeps publishing original research that helps other voter mobilization groups improve their programs. If you know someone who works at VoteAmerica, or if you’re a volunteer, mention that! People love a personal connection.
• Fourth, remind people that the clock is ticking! Tell them exactly how many days there are until Election Day.
• Finally, ask if anyone has any questions or is ready to make a pledge. (THEN WINK AT YOUR PLANTED DONORS)

The Debra Cleaver guide to throwing an epic friendraiser was originally published in VoteAmerica on Medium, where people are continuing the conversation by highlighting and responding to this story.

President Trump continues his attempts to delegitimize the election months before it takes place. First he launched a sustained attack against voting by mail, then he suggested delaying the election. He’s even placed a campaign donor with clear conflicts of interest as Postmaster General of the United States Postal Service. Now, the President has explicitly told reporters that he’s blocking $25 billion in emergency funds for the U.S. Postal Service because Democrats want to expand mail-in voting during the coronavirus pandemic.

It’s like something out of a bad episode of House of Cards — a president attempting to crash the Postal Service — an iconic American institution with roots older than the country itself — in the middle of a pandemic in order to block the one way that people can vote safely in November without putting their lives at risk. You cannot make this stuff up.

According to the Washington Post:

“Now, they need that money in order to make the post office work, so it can take all of these millions and millions of ballots,” Trump said in an interview with Fox Business Network’s Maria Bartiromo. He added: “Now, if we don’t make a deal, that means they don’t get the money. That means they can’t have universal mail-in voting, they just can’t have it.”

Unbelievable. The U.S. Postal Service is enshrined in the Constitution and is a public service to the people of this country. It is not a business that needs to turn a profit — it is a government service. And during COVID, it’s a lifeline for many Americans — including Veterans and the elderly — that depend on the mail service for things like life-supporting prescription medicines.

Yet, it appears that President Trump’s hand-picked Postmaster General, Louis DeJoy, is implementing his boss’s plan to dismantle the Post Service and thwart mail in voting.

DeJoy, a highly partisan friend of and donor to the president, is using his newfound power as Postmaster General of the USPS to deliberately disrupt and delay the delivery of mail ahead of the November election. This means millions of mail-in ballots are at risk of not being submitted in time. We need to work that much harder now to make sure everyone requests their mail-in ballots now and returns it well ahead of the election.

In the midst of the COVID pandemic, as Americans seek to vote by mail for personal safety and public health concerns, DeJoy is actively dismantling the USPS from within. He’s purposely slowing down mail service and removing mailboxes in local communities, and removing critical mail sorting equipment with no explanation — the exact sorting equipment that sorts election mail. He’s freezing hiring and is trying to force non-union postal employees into early retirement. And DeJoy is trying to nearly triple the cost of sending mail-in ballots this November. This is not by accident — it is by design and is part of an ongoing plan to disrupt the November elections by causing many mail-in ballots to miss deadlines and votes not be counted.

Instead of making sure that all Americans can safely vote in November during a pandemic, and that the Postal Service has the resources to meet the challenge of an influx of people voting by mail for the first time due to COVID, President Trump is blocking funding for and smearing the Postal Service. He continues to spread lies about vote by mail suggesting with zero evidence that they’ll cause widespread fraud. And he’s brazenly admitting this is all to hurt Democrats and improve his own chances of reelection.

The truth is that mail-in ballots are secure — they include barcodes and signature matches, leave a paper trail, and — unlike electronic voting machines — paper ballots are not hackable. The President and multiple members of his Administration vote by mail. His attacks on vote by mail are untrue. His attacks on the Postal Service are unconscionable.

Combined, his actions are shocking and a threat to our elections — and our democracy.

By attacking the Postal Service and attempting to thwart safe and secure vote by mail, the president is forcing Americans to risk their health and that of their communities by standing in hours-long lines to cast their rightful vote in November — or not vote at all.

There’s no time to waste in fighting this blatant effort to undermine the election. Slowing mail service in America greatly reduces the amount of time people have to vote. VoteAmerica is working to supercharge our efforts to sign-up millions of citizens to vote by mail NOW and push them to get their ballots in before partisan efforts to destroy the United States Postal Service prevent it. We need your help.

We cannot let DeJoy or the Trump Administration kill the Postal Service. In addition to eliminating hundreds of thousands of jobs, further dampening an economy already in distress, it would rob every American — Republican, Democrat, Independent — of an essential lifeline to voting during a national health crisis. We will not stand for that.

Join us in standing up for our Postal Service. Support our work and donate to VoteAmerica today.

Thank you,

Debra Cleaver
Founder & CEO
VoteAmerica.com

The President Just Admitted He’s Blocking Funds for the U.S. Postal Service to Stop Mail-In Voting was originally published in VoteAmerica on Medium, where people are continuing the conversation by highlighting and responding to this story.

NB: these posts start out as emails that go to our VIP email list. That list consists our donors, partners, and generally smart and interesting people who we know in real life. The recipients have asked me to cross-post those emails somewhere public, so they can share them with their networks. Medium is public, and the emails aren’t confidential, so I’m happy to oblige.

I’m back with another kind of long, but very important VoteAmerica email for you. I swear I don’t get paid by the word, I just really love talking about the smart things we do.

According to the Independent Voter Registration report of 2018, 100 nonprofit organizations spent approximately $300 million dollars to register 7 million voters. Partisan and dark-money groups spent exponentially more. This is a tremendous amount of money for a relatively modest return, with the vast majority of that money spent on staff who are paid to do in-person, paper based voter registration. In-person, paper based voter registration is the same tactic that I used when I ran my first voter registration drive on campus in 1996. I was able to register dozens of people a day using this tactic, and I was quite pleased.

25 years later, we’re still reliant on the same tactics I used as a college student. This is baffling. Voter registration in the United States is expensive because the dominant registration tactic is older than the internet. Face-to-face, in person voter registration is slow and horribly expensive. We can do better.

In 2020, VoteAmerica ran a pilot program with CreditKarma, in which CreditKarma leveraged our API to integrate our tools into their existing online workflow. In less than three months this partnership drove nearly 600,000 Americans through our tools where they were able to register to vote, check their registration status, request a vote-by-mail ballot, find their polling location, and more. By capitalizing on CreditKarma’s existing network of users, we were able to reach those nearly 600,000 voters in an incredibly cost-effective way.

This is the future of large-scale voter registration and mobilization programs and we’re excited to lead the way. As we plan for the midterm elections, we’re going to continue to grow our strategic partnerships with companies that already have millions of existing users. We’re focusing specifically on companies that exist to help you fill out paperwork, such as your annual taxes and your change of address forms. We’ll also be focusing on companies that help college students register for classes online. This three-pronged focus will help us reach literally tens of millions of Americans for the cost of a handful of software developers. The amount of time and money it takes to build out company-specific integrations is incredibly small when compared to the time and money invested in in-person voter registration.

This new model is truly meeting voters where they are, at scale, and at reduced cost. These partnerships will enable us to register and mobilize millions of voters without spending tens of millions of dollars.

Now it’s time for me to ask you to get involved. VoteAmerica is the already the most-cost effective voter turnout organization in America. We’re smart, and we’re frugal, and we’ve been at the forefront of driving costs down for decades now via our various past initiatives (including Vote.org). This strategy will drive our already low costs even lower, while providing an exceptional return on investment. We’ll use every dollar we receive to expand our team and run the programs (like these partnerships) that will yield the highest impact.

If you’re able, please consider a tax-deductible donation today.

• You can donate online.
• You can donate stock.
• You can donate crypto.
• Your company might even match your donation!

One last thought: donating stock is one of the best ways to support our work. You get to write off the full value of the stock, without having to liquidate it and pay capital gains taxes. It’s a win for you tax wise, and a win for VoteAmerica (or any other charity) because we receive the full value of the stock.

Let’s go register a ton of voters.
Debra

We are overpaying for voter registration. Let’s fix that. was originally published in VoteAmerica on Medium, where people are continuing the conversation by highlighting and responding to this story.

I send an email to the VoteAmerica VIP list about once a month. This is a relatively small email list that reaches a few thousand people. Folks on that list have asked me to cross-post the emails on Medium so they can share with their friends. This email has been very lightly edited for a wider audience. It was originally sent on April 8, 2021.

Hey friends.

First, I know it’s been a while since I’ve sent one of these emails. Generally if i’m not talking (or emailing) it’s because i’m thinking. Mostly I’ve been thinking about the very, very long list that I wrote in January 2020 of things that could go wrong with the election, and how I legitimately did not think to put “armed insurrection” on that list. Second, I fully intended to send one of these emails on Jan 6, 2021, but scrapped the email when I woke up to an armed insurrection. Third, by “these emails” I mean specifically our VIP emails, which are very different in tone and content than the emails we send to our 3 million person general list. If you’re receiving this email, it’s because you’re tagged as a VIP in our database. If you don’t want to receive these long and somewhat serious emails from me in the future, please click the unsubscribe link in the footer.

Now that we’ve gotten the housekeeping out of the way, let me catch you up on what’s going on in general. Republican lawmakers in 47 states have put forth 300+ voter suppression bills so far this year. They claim that these bills are a response to “voter fraud.” This is untrue. These bills are a response to record breaking voter turnout, specifically among marginalized communities. The Republican platform is increasingly unpopular with the majority of voters, which makes it harder for Republicans to win elections. So these bills serve a few purposes: they’ll make it harder for Black and Brown people to vote; they’ll make it more likely that politicians with unpopular platforms will win; and they’ll make it damn near impossible to remove incompetent, corrupt, and sometimes openly evil lawmakers from office.

Fantastic.

It actually gets worse. If you’re wondering how exactly lawmakers could introduce 100s of bills so quickly, it’s because they’re all written by the American Legislative Exchange Council (ALEC), which is a sort of comically evil “think tank” funded by Charles Koch. Charles Koch is the last of the Koch brothers, arguably the smartest, and one of the primary funders of the modern “conservative” movement, including the Tea Party. He’s also a billionaire who made his money in coal. And he’s brilliant: instead of running for office himself, he just buys lawmakers, drafts their legislation, and lets them execute on his plans. And right now his plans involve suppressing voter turnout.

Which brings me to Georgia. If you’re on this list, you read the news, so you know at least the broad strokes of what’s going on in Georgia. Republican lawmakers in Georgia are pretty clear that they do not care for high turnout rates among Black folks, so they’ve passed Senate Bill 202. Senate Bill 202 includes new burdensome ID requirements for absentee voting, reduces the number of drop-boxes for absentee ballots, makes it illegal for groups to send pre-filled absentee ballot applications to potential voters, and makes it a misdemeanor to give food or water to voters waiting in line to cast ballots.

This bill is both brilliant and evil. Brilliant because the whole “it will be illegal to hand out food or water to voters waiting in line” is a red herring. It’s outrageous, and it will almost definitely be overturned in the courts. In the meantime, it serves to distract us all from the actual goal of this bill: preventing the charities that register voters and help them sign up for absentee ballots from doing their work. This bill has nothing to do with what happens once you’re lined up to vote and everything to do with making sure that citizens simply can’t vote.

So now we’re suing the state of Georgia.

VoteAmerica is the lead plaintiff in a lawsuit filed yesterday: VoteAmerica v. Raffensperger. (Raffensperger is the Georgia Secretary of State.) You can read the lawsuit here: https://www.democracydocket.com/wp-content/uploads/sites/45/2021/04/VoteAmerica-Complaint.pdf

We’re focused specifically on the section of the bill that makes it illegal to send “pre-filled absentee ballot applications” to voters. Pre-filling absentee ballot applications is literally what one of our tools does: voters enter their information in our web-based tool, we transfer that information to the correct form, and then we provide the correct mailing address and other directions. If a voter doesn’t have a printer, we’ll also print and mail this form for them. In 2020 alone, we sent 48,000 pre-filled absentee ballot applications to voters in Georgia. These are folks who likely would not have voted without our assistance.

So that’s what’s going on in Georgia. This bill has nothing to do with food and water, and everything to do with attacking the charities that help people vote. And they’re not just attacking us and our co-plaintiffs: they plan to bankrupt us. Right now we’re looking at a $100 fine for each one of these forms that we send. Cute.

Along with our co-plaintiffs and our lawyers at the Campaign Legal Center, we’re going to do everything in our power to fight this ridiculous law. And we’re moving quickly, because we want this insidious law overturned in 2021, not 2022. So we’re bringing these issues to the forefront now, in 2021, as far ahead of the midterm elections as possible. Midterm elections are arguably more important than presidential elections, and we don’t have any time to waste.

We also don’t have a lot of money in the bank right now. At some point late in 2020, we decided to leave it all on the field, and by all, I mean almost all of our money. We decided that increasing turnout was more important than job security, and that we would rather be unemployed than watch our country descend further into fascism. So we didn’t finish the year with a financial buffer, but we did finish 2020 with a mostly intact democracy.

So now we’re building up our war-chest again, starting with funding for this battle with Georgia. If Georgia gets away with this, other states will follow suit, and they won’t stop with absentee ballot applications: I am 100% certain that the next onerous laws will make it illegal for charities to send out pre-filled voter registration forms. Which will be devastating for democracy.

In the meantime, if you’re able, please consider a tax-deductible donation today. We’re smart, we’re aggressive, and our programs are cost-effective.

• You can give online at this link: https://secure.actblue.com/donate/voteamerica-2021-georgia-lawsuit
• If you plan to give more than $1,000, this document has details on wire transfers, DAFs, and stock donations. https://docs.google.com/document/d/1dpv0GLPRYPjA7zCw-zNyQU8-PdrorsfIV31Bh6AREI0/
• Speaking of donating stocks: this is probably one of the better ways for you to donate. You get to write off the full value of the stock, without having to liquidate it and pay capital gains taxes. It’s a win for you taxwise, and a win for VoteAmerica (or any other charity) because we receive the full value of the stock.

Thank you,

Debra

Here’s some early press coverage on the lawsuit.

• https://abcnews.go.com/Politics/civic-groups-file-lawsuit-challenging-absentee-ballot-provisions/story?id=76945055
• https://www.huffpost.com/entry/voting-groups-sue-georgia-law-absentee-restrictions_n_606e51c3c5b6034a70847ae3

VoteAmerica is suing Georgia. was originally published in VoteAmerica on Medium, where people are continuing the conversation by highlighting and responding to this story.

Americans don’t need to be convinced to vote. They need to be able to vote. It’s harder to vote in the United States than in any other nation with democratically-elected leadership. We know that voters turn out when you ask them to, and when they have the resources to be able to. But some states have systematically made it hard to register and to request a mail-in ballot (also called an absentee ballot), and it’s no accident that these barriers disproportionally affect low-income voters and voters of color.

Partisan campaigns focus on what are called “high-propensity” voters — people who usually vote. This creates a self-reinforcing system where you have a group of voters that face more barriers getting to the ballot box, and then these voters get classified as “low-propensity” and nobody reaches out to help them.

At VoteAmerica, we break this cycle by identifying these barriers, building tools to remove them, and then proactively reaching out to voters (tens of millions of them in the run-up to the 2020 General Election!) to help them find and use these tools so that everyone can make their voice heard and have their vote counted.

Requesting a ballot

One of these key barriers is being able to request a mail-in ballot. Even in the midst of a pandemic, many states do not have an online system for requesting a ballot. So voters need to find the ballot request form, print it, sign it, and then mail it or hand-deliver it to their election office. And with tens of millions of Americans out of work due to the pandemic, even fewer people have access to a printer: home printer ownership is in the low single digits and has been for years.

But in about ten states — comprising tens of millions of voters — you’re allowed to submit the ballot request form by email or fax, as long as it still has a “wet” signature (a real pen-on-paper signature that’s been scanned in). But to do that, you’d need to either still have a printer and scanner, or have the technical know-how to use something like Apple Preview to scan in your wet signature and apply it to the application form PDF. And in states that allow fax but not email (New Hampshire and Kentucky), you’d also need access to a fax machine!

So at VoteAmerica, we built an e-signature workflow that let you complete your ballot request form — including a wet, pen-on-paper signature — entirely online, from a web browser on your phone or computer.

The VoteAmerica ballot request workflow

First, the user enters in their basic biographical information. This is used both to fill out the form, and so we can follow up with them later to help them complete and return their ballot once they receive it.

Then, the user enters in a set of state-specific information. Unlike voter registration, where anyone from any state can use the National Voter Registration Application, each state has a different form for requesting a mail-in ballot, each with different fields to fill out and questions to answer (well, that’s not quite true — every state accepts the Federal Post Card Application for overseas and military voters, but no state accepts that form for domestic voters). So we reviewed and digitized all 51 forms — for every state and DC — to build state-specific forms. Massachusetts, for example, keeps it pretty simple:

Then, we ask the user to take a picture of their wet, pen-on-paper signature. On mobile, we provide a camera UI right in the page, and on desktop, we let you text the picture to us (we don’t use a camera UI on desktop because it’s a bit awkward to try and hold a paper with your signature up to your webcam without the paper blocking your view of your computer screen).

As soon as the user takes the picture or texts it to us, we show it and provide an opportunity to crop and rotate:

Then, we process the image to extract a clean, black-and-white signature that we can apply to the form:

And that’s it! The user has filled out the whole form using our web-based tool. We then send the form directly to the user’s Local Election Office, either via email or fax (depending on the state, and what that state and Local Election Office will accept). The election office gets a nice, clean form like this:

Electronic vs. digital signatures

We’re taking advantage of an important distinction here between an electronic signature and a digital signature. You’ve probably used a digital signature before — it’s a signature that’s never existed in physical form, like your finger on a touchscreen or typing in your name and having a computer show it in a cursive font. Some states, like Massachusetts, do accept digital signatures, but they’re not accepted in most states. So instead, what we’re doing here is an electronic signature: a real pen-on-paper signature that we’re just transmitting via technical means. Think of it like Mike Teavee from Charlie and the Chocolate Factory, who gets transported via television: we’re taking a real, physical signature, and we’re just applying it to the document and sending it electronically.

Electronic signatures are valid everywhere: the laws governing them are federal laws enacted to establish that faxed contracts are just as binding as mailed contracts.

Even in states where we could use a digital signature, we use electronic signatures. One of the most common causes for mail-in ballots to be rejected is because the signature on the ballot doesn’t match the one the government has on file from the voter registration or ballot application, and electronic signatures are much more likely to look like the voter’s signature on file than a digital signature.

So where does computer vision come in?

You might have noticed in that walk-through that there’s a pretty interesting step where we go from the signature in the photo to the clean, black-and-white signature:

How do we do this? We stitched together a number of image processing and computer vision techniques to reliably extract signatures from user-captured images.

Why is this hard?

If you’re new to the world of image processing and computer vision, you might be wondering what’s so hard about this. After all, when you look at that photo, you immediately see the signature — the human brain has no trouble at all identifying what part of that image is the signature, and what part is the background. Computers, however, have a much harder time figuring that out. To understand why, take a look at this image: which square is lighter: square A or square B?

You brain knows immediately that square A is darker and square B is lighter. However, those two squares are actually the exact same color:

(I’ve just added two colored bars here to make it clear that the squares are the same color; nothing else about the image changed).

When you look at this image, you know about things like shadows and checkerboards, and so your brain automatically compensates for the shadow cast by the cylinder — so even though the actual pixels of the image are the same color, your brain knows that square A is a dark square, and square B is a light square with a shadow on it. But it’s not so easy to teach a computer about this.

And this is pretty much the problem we’re trying to solve. When the user gives us a picture of their signature, we need to tell the difference between the dark parts of the image — the actual ink of the signature — and the parts that are just the paper, with stray marks or shadows.

There’s a couple of very simple techniques to do this that don’t work very well. The simplest is a basic threshold: we look at each pixel in the image, convert it to grayscale (just a shade of gray, with no color, like a black-and-white photograph), and then we see whether that pixel is closer to white, or closer to black. This works pretty well for really clear, simple images, but fails for lots of real-world cases like shadows and blurry images:

Another common technique that works a little better is Otsu Binarization. This technique works pretty similarly to the simple threshold: it looks at each grayscale pixel, and checks whether it’s closer to black or to white. But instead of dividing down the middle — more than 50% black ends up as black and less than 50% black ends up as white — it picks a smarter threshold based on the whole image to maximize how much detail is preserved. This works a bit better, but still not great:

Our algorithm, on the other hand, uses a number of different techniques to achieve consistent results:

As you can see, this works in a wide variety of cases on lots of different input images.

Our algorithm

Our algorithm uses six steps, which I’ll go through in some detail:

The six steps are:

• Grayscale: convert the image to grayscale
• Denoise: remove background noise with non-local means denoising
• Threshold: apply adaptive thresholding to convert the image to black-and-white
• Open: remove stray marks
• Blur: soften edges
• Crop: identify the core image and remove the rest

Step one: grayscale

The grayscale step is pretty straightforward: we take each pixel, average the red, green, and blue components of that pixel, and then use that as the color of the pixel. This strips away all of the color information and makes every pixel a shade of gray.

This step simplifies everything that comes after: because we don’t care about the color information, discarding it gives us a much simpler input.

Step two: denoise

Now things start to get interesting. We want to remove background noise from the image — the random grey speckles that show up in digital images from cameras. This makes it easier for the rest of the algorithm to focus on the important parts of the image, rather than the random background noise.

Background noise is just a random variation in brightness and color that’s caused by the image sensor in a digital camera. It’s particularly noticeable in low-light photographs, which is pretty common when you’re taking a picture of your signature indoors — that’s part of the reason photographers use lots of lights when shooting professional photos. You can think of it as a random speckle pattern overlaid on top of the image:

If we have multiple copies of the source image, with different noise (for example, if we had multiple frames of a video — each frame showing the same thing, but with different noise patterns) we can take advantage of the fact that adding together noise gives us a consistent gray:

So if we add together multiple copies of the source image, each with different noise, the noise averages out and we get a de-noised image:

This works great! Except that in our case, we only have a single copy of the user’s signature. So we use a technique called non-local means denoising. With this technique, we find parts of the image that look similar, and then average those out to remove noise. So for example, the red and yellow boxes in this image:

And then by doing that over and over — finding lots of pieces of the image that look similar and averaging them out — we can de-noise the whole image:

Step 3: Adaptive thresholding

Adaptive thresholding is the meat of the algorithm: it’s what removes the background and converts the whole image to black-and-white. And it works much better than the simple thresholding and Otsu Binarization we talked about above, because of a key insight: we can’t apply the same threshold to the whole image.

Instead of trying to pick a threshold between black and white, and use that for the whole image, we instead pick a different threshold for each small part of the image, like this:

For each pixel, we look at an area around it to determine what qualifies as “light” or “dark” in that part of the image, and then use that local threshold to determine whether the pixel should be black or white in the output. This means that we’re not comparing the shadowy part on the right side of this image directly to the brighter part on the left side of the image.

Step 4: Opening

While this image has gone pretty smoothly so far — a nice, clean signature with no stray marks — sometimes a bit of dirt or a stray dot of ink still makes it through. So we use a technique called morphological transformation, which is a family of operations based on the shapes in an image. In particular we use a technique called opening, which is a combination of erosion and dilation.

Erosion takes an image and shrinks all the shapes in it:

And dilation does the opposite:

Doing one and then the other — eroding and then dilation — has a somewhat magical effect: most shapes are left the same (because the two operations cancel each other out), but small dots and marks are removed (because the erosion totally removes them, so there’s nothing left to dilate).

This combination of erosion and dilation is called opening, and it’s what we do to the processed signature images. It doesn’t have a big impact on this signature that we’re working with now, but as you can see from above, it helps to remove small marks in many cases.

Step 5: Blur

One more quick step before we do our final cropping: we apply a slight blur to the image. This makes the image look a bit more natural — our adaptive thresholding converted the image to pure black-and-white with sharp edges, but physical writing doesn’t have such sharp edges. The ink bleeds a bit into the paper, giving softer lines. By applying a subtle blur, we make the image look a bit more natural and more legible. We use a gaussian blur, which means that for each pixel, we take a bit of the color of the pixels around it (weighting closer pixels more heavily), and use that as the new color of the pixel. So pixels right on the edge of black and white end up grey, because they take on some of the color of the pixels around them.

Here’s what that looks like zoomed in:

Step 6: Cropping

Our final step is to crop the image. This is actually really important: we’re going to be placing these signatures on PDFs, and those PDFs often don’t leave much space for the signature. So it’s important that we crop the image as tightly as possible, so we can use all of the available space for the signature, and not waste any space on empty background. We’ve asked the user to crop the image, but users mostly use that opportunity to remove other objects from the photo, rather than carefully cropping the image as tightly as possible.

To crop the image, we first segment it into connected shapes. For each of these shapes, we apply a minimum threshold to determine whether it’s big enough to care about. In this example, we have 4 shapes, of which 3 are big enough to care about:

We then calculate the tightest possible crop that includes all of the large shapes, and use that:

Note that we don’t actively remove the small shapes (the ones in red) — if they’re within the crop, they get to stay. This lets us remove stray marks around the edge of the image, without removing things like the dots on top of i’s or j’s.

Final result

And that’s it! A robust pipeline to extract signatures from a wide variety of user-captured images.

Wrap-Up And Learnings

This algorithm works well for a wide variety of signatures — and that part is really important to us. At VoteAmerica, we work to turn out underrepresented voters, and so we need to be really careful that our technology doesn’t unintentionally exclude certain kinds of voters.

Here’s an example of a signature we do pretty well at now, but some earlier versions of the algorithm didn’t do so well at.

Before we tuned it just right, the algorithm often removed the marks and dots that make this signature distinctive. We were overly reliant on the idea that a signature is one or two connected pieces, but that’s not true of a lot of people’s signatures.

In particular, a lot of younger people, and lower-income people who often don’t sign as many things, don’t have a smooth, simple signature. And so if our algorithm didn’t work well for those kinds of voters, we’d run the risk of a signature mismatch between the voter’s actual signature, and our electronic scan of that signature.

And so one of the key design constraints of our algorithm was that when it doesn’t work — like in this example, where it’s picked up the lines from the paper — it should be really clear to the voter why it didn’t work, so they can look at the result and try again.

We prototyped some versions that used more advanced machine learning systems to try and segment out what part of the image was a signature and what wasn’t, by training a model on a bunch of signatures. But the problem is that when that fails, it’s not obvious — it just picks the wrong part of the image. And that’s much harder for someone to understand and fix.

So even though we’ve built a system that’s much too complicated for most voters to understand, we’ve built a system where you can understand the output, and have an intuitive sense of what failed and how to fix it even if you don’t know how the underlying algorithm works.

Complex systems don’t always have to have complex results.

Check out the code

All of this code — for the algorithm itself, and the AWS Lambda deployment that we use to run it and handle things like receiving signature images via MMS — is available on our GitHub account at https://github.com/vote/signature_extractor.

The computer vision tech that helped 15,000 voters request a ballot online was originally published in VoteAmerica on Medium, where people are continuing the conversation by highlighting and responding to this story.

(This story was originally posted on the Twilio Blog)

At VoteAmerica, we work to help Americans register to vote, and request vote-by-mail (also known as absentee) ballots. We use technology to remove barriers to voting and make sure that everyone can vote safely, easily, and securely.

The 2020 election will be like no other: with a global pandemic, we anticipate a massive increase in the number of Americans who will vote by mail this November, and we’re making sure that requesting a mail-in ballot is as accessible as possible. Requesting a ballot can be difficult: in many states, you need to print a form, sign it, and mail it — and very few households have access to a printer, especially with so many people working from home. In 19 states, voters can request a ballot online via an official state system. And 11 more states — covering over 50 million voters — allow you to submit a ballot application via email or fax.

And it’s those 11 states that we’ll focus on in this post: states where you can submit your application electronically… if you have the technical know-how to scan your signature onto the paperwork and email it or fax it. And in a few of these states (like Kansas and New Hampshire), email isn’t an option: you have to use a fax machine (and who has access to a fax machine at home any more?). Even in states that typically allow you to email your application, some counties have been so overwhelmed by the volume of email that they’ve temporarily stopped accepting emailed applications.

So at VoteAmerica, we set out to build a system that makes it easy for voters to fax their ballot applications, to help voters in states that don’t accept email or counties where the email systems aren’t working. And that’s where we ran into a tough technical problem: how to reliably send faxes. We have to be very, very sure that our faxes are going through: a lost application means that a voter could be waiting for a ballot that never comes. Twilio offers a simple, easy API for sending faxes — so what’s the tough part?

How Faxes Work (And Don’t Work)

Fax machines work via the telephone network. When you send a fax, you’re actually making a phone call — but rather than connecting two people to each other to talk, you’re connecting two fax machines. The sending fax machine scans the document and encodes it for transmission over the phone line, and the receiving fax machine decodes the document and prints it.

So unlike email, there’s a whole lot of things that can go wrong. Most commonly, the phone line is busy! If you’re using a web-based fax receiver (like Twilio), this isn’t a problem: Twilio handles the complex phone network magic to be able to receive multiple faxes simultaneously at the same number. But we’re not sending faxes to Twilio — we’re sending them to election offices, most of which are still using an old-fashioned fax machine connected to a landline. And just like phone calls, only one fax can be using the line at a time. So if someone else is trying to send a fax to that election office, we’ll get a busy signal when we try to send the voter’s application.

Even if the line isn’t busy, there’s still other things that can go wrong. Fax machines typically have limited memory (they’re just storing documents long enough to print them), so if the machine is out of paper or ink and runs out of space to store incoming documents, it won’t accept any more incoming faxes. Even simpler, the fax machine could be off or unplugged.

All this to say: we need to be really careful about sending faxes, and anticipate that a lot of our attempts to send a document won’t work. While Twilio’s API doesn’t completely solve this out-of-the-box — sending a document via the Twilio Fax API just tries to send once and fails if the line is busy or disconnected — it does give us all the building blocks we need to build a reliable fax system.

The Fax Gateway

To ensure our faxes get sent successfully every time, VoteAmerica built an open-source Fax Gateway: a system that sits on top of the Twilio API to handle queuing and retrying faxes. To make this system cheap to operate and easy to maintain, we built it using Amazon SQS, AWS Lambda, and the Serverless Framework.

The fax gateway handles a couple of problems:

• If a fax doesn’t go through, it automatically queues it up to try again after a few minutes. It tracks how many times we’ve tried to send a particular fax, so if it fails too many times we can escalate that problem to an engineer to investigate and let the voter know their application wasn’t sent successfully.
• It queues the faxes so we’re never sending more than one fax at a time to a particular number. Because most election offices are using a traditional physical fax machine that can only receive one document at a time, we don’t want to try sending multiple documents at the same time — we’d end up competing with ourselves and getting a lot more busy signals.

The basic flow is that a pending fax is written to an SQS queue. This is a FIFO (first-in, first-out) queue. SQS FIFO queues have a number of extra features that help us out:

• They can deduplicate messages. This means that if the API call to write a message to the queue fails, we can safely retry it because the queue will remove the duplicate if it’s written twice.
• We can group messages via a “message group ID”. This partitions the queue — basically, we get a separate logical queue for each message group ID. Messages within a group are processed strictly in order, and only one message from each group is processed at a time.

This second feature — the message group IDs — is what lets us make sure we don’t jam up a particular election office by sending multiple faxes at once. We use the destination fax number as a message group ID, and SQS will make sure that only one message from each group (meaning one fax for each destination number) is being processed at a time.

The Life Of A Fax In The Fax Gateway

When we want to send a fax, an application writes it to the Fax Queue. From there, it’s sent (via the Twilio API) to the Fax Processor. The fax processor sends the fax and checks if it went through. If it did, the fax gateway writes a message to the Webhook Queue, which sends a notification back to the application. If the fax didn’t go through successfully, we write the message to a Retry Queue, which holds on to it for a few minutes, and then sends it back to the Fax Queue to try again.

When an application wants to send a fax, it just writes it to the Fax Queue. Here is how this can be done from Python, using the boto3 library:

import boto3
sqs_client = boto3.client('sqs')

payload = json.dumps({
  "fax_id": str(fax_uuid), # A unique ID for the fax
  "to": str(fax_to), # A E.164-formatted phone number
  "pdf_url": fax_pdf_url, # We use a presigned s3 GET URL
  "callback_url": callback_url, # A URL in the application that
                                # receives webhooks
    }
)

sqs_client.send_message(
    QueueUrl=FAX_GATEWAY_SQS_QUEUE,
    MessageBody=payload,
    MessageGroupId=fax_to,
)

And then the Fax Gateway’s queues and Lambda functions take over. Let’s take a deeper look at how one of these functions — the Fax Processor — works.

The Fax Processor

The Fax Processor is the heart of the Fax Gateway. It reads the faxes from the Fax Queue and sends them via the Twilio Programmable Fax API:

# Send the fax to Twilio
twilio_fax = client.fax.faxes.create(
    from_=PHONE_NUMBER, # A phone number from our Twilio account
    to=fax_record.to,
    media_url=fax_record.pdf_url,

    # Our faxes can contain PII -- instruct Twilio to not retain
    # a copy of the PDF.
    store_media=False,

    # Fail fast (after 5 minutes) rather than leaving the fax
    # queued -- we have our own retry logic, and this lambda
    # function times out after 15 minutes so we'd rather
    # gracefully handle the failure ourselves rather
    # than have Twilio hold it in their queue for a long time.
    ttl=5,
)

One important design decision was how to handle failed faxes. When you send a fax via Twilio, you generally provide a status callback. Twilio will try to send your fax, and let you know via the status callback whether it went through. So we could have the Fax Processor fire off the fax via Twilio, set the status callback to invoke another Lambda function, and then return (so Twilio would send the fax in the background, after the Fax Processor returns). However, this would mess up all the work we’ve done with message groups: we’re trying to only send one fax at a time to a destination number. So it’s important that the Lambda function doesn’t return until the fax is done sending.

So instead of using the Twilio status callback, we have our Lambda function stick around, and poll the Twilio API. This keeps the Lambda function running until the fax is done sending (successfully or not), so we maintain this “lock” on the destination number:

# Twilio fax status codes indicating still-in-progress or success -- all other
# codes are considered failures.
#
# https://www.twilio.com/docs/fax/api/fax-resource
TWILIO_STATUS_PENDING = ("queued", "processing", "sending")

# How frequently to poll for fax status, in seconds
TWILIO_POLL_INTERVAL = 15

def poll_until_fax_delivered(fax_sid: str) -> Any:
    while True:
        try:
            fax = client.fax.faxes(fax_sid).fetch()
        except Exception as e:
            # If there was an error getting the fax status, just
            # log it and keep polling -- we don't want to let a
            # transient error cause the whole lambda function to
            # fail.
            logging.exception("Error while polling for fax status")

        if fax.status not in TWILIO_STATUS_PENDING:
            return fax.status

        print(f"Fax has pending status: {fax.status}, waiting")
        time.sleep(TWILIO_POLL_INTERVAL)

We also distinguish between a failure to send the fax due to a busy signal or other expected problem — the stuff that we expect to happen every once in a while even if everything is working correctly on our end — and a failure to send the fax due to a problem with AWS, Lambda, or Twilio. If we encounter an unexpected error (for example, if the Twilio API returns a 500 error code), we use the normal Lambda error handling mechanism: we throw an error from the Lambda function, and the message will end up back in the Fax Queue to try again after the SQS visibility timeout. If we get an expected error — Twilio tells us the fax line was busy, for example — we don’t return an error from the Lambda function. Instead, we write the message to the Retry Queue and return a success.

The Retry Processor

Messages sit in the Retry Queue if we’ve tried to send the fax, but it didn’t go through. Usually, this is because the fax line is busy, so we want to wait a bit for the fax line to clear up before we try again. The Retry Queue is a delay queue, which means that we’ve configured it to hold onto messages for about 15 minutes before we deliver them. This lets us space out our retries, so the receiving fax machine has time to stop being busy, or have its paper or ink replaced, before we try again.

We keep the Retry Queue separate from the Fax Queue so that while we’re waiting to re-send that fax, we can try other faxes to that number (so if the fax is unprocessable for some reason — say, for example, the PDF is corrupt or invalid so Twilio can’t send it — it doesn’t clog up all messages being sent to that number).

The Retry Processor itself is very simple: it just reads the messages, and then writes them back to the Fax Queue:

# Take a fax record from the retry queue and sends it back to the
# fax queue to be retried
def handler(event: Any, context: Any) -> Any:
    # We set batchSize to 1 so there should always be 1 record
    assert len(event["Records"]) == 1
    record = event["Records"][0]

    fax_record = Fax.json_loads(record["body"])
    enqueue_fax(fax_record)

The Webhook Processor

If we do successfully send the fax, or we’ve failed to send the fax a whole lot of times (we typically attempt to send a fax 20 times before giving up), we write a message to the Webhook Queue, which delivers a notification (via a POST HTTP request) back to the application that was trying to send the fax.

The code for the Webhook Processor is also pretty simple — we just load up the message from the Webhook Queue and deliver it via an HTTP request:

# Take a webhook record from the queue and try to send it to the calling
# application
def handler(event: Any, context: Any) -> Any:
    # We set batchSize to 1 so there should always be 1 record
    assert len(event["Records"]) == 1
    record = event["Records"][0]

    # Webhook is a Python dataclass that corresponds with the
    # message that we write to the Webhook Queue
    webhook = Webhook.json_loads(record["body"])

    response = requests.post(
        webhook.callback_url,
        data=webhook.payload.json_dumps(),
        headers={"Content-Type": "application/json"},
    )

    response.raise_for_status()

One important distinction here between the Retry Process and the Fax Processor is that the Retry Processor uses the typical SQS/Lambda error handling: rather than have a separate retry queue for failed messages, we simply raise a Python exception. The SQS/Lambda integration uses this as a signal that the processor failed, and the message should be put back into the Webhook Queue to be sent again after the SQS visibility timeout. So if the application that’s supposed to handle this webhook fails and returns a 500, we’ll just end up processing this message again and re-send the webhook. We configure an SQS dead-letter queue so if we’ve tried to deliver the webhook a whole bunch of times, we’ll eventually give up and retain the message in the dead-letter queue for us to inspect and debug.

Using the Fax Gateway

We’ve released the Fax Gateway as an open-source project so anyone can use it to send faxes reliably. It’s easy to deploy in your own AWS account using the Serverless Framework.

To deploy the Fax Gateway, you’ll first want to fork the repository so you can make your own changes to the Serverless configuration. You’ll want to read through the configuration file to make sure it’s set up correctly for your AWS environment. In particular you’ll need to customize a few things:

• We deploy our fax gateway in the us-west-2 region. If you use a different region, you’ll want to change the line that says region: us-west-2 to refer to your AWS region.
• We deploy the Fax Gateway in our VPC, so we can use AWS security groups to ensure that only the Fax Gateway can send requests to our webhook endpoint. You’ll want to make sure that the vpc section of the configuration has the right configuration for your VPC, or remove that section altogether if you don’t want to deploy the Fax Gateway into a VPC.
• We use DataDog to collect logs and metrics, so we’ve set up the Fax Gateway to deliver logs via the Datadog Forwarder. If you don’t use DataDog or the Datadog Forwarder, you’ll want to remove the serverless-plugin-datadog plugin.
• We use Sentry to track errors. If you don’t use Sentry, remove the SENTRY_DSN and SENTRY_ENVIRONMENT environment variables.

We use AWS SSM to store the credentials for the Fax Gateway. You’ll need to configure the following SSM Parameters:

• fax_gateway.common.sentry_dsn: If you use Sentry, set this to your Sentry DSN to send Fax Gateway errors to Sentry.
• fax_gateway.common.twilio_sid and fax_gateway.common.twilio_auth_token: Set these to your Twilio credentials so the Fax Gateway can authenticate with Twilio to send faxes.
• fax_gateway.prod.twilio_phone_number: Set this to the phone number in your Twilio account that the faxes should come from. If you plan to use multiple environments (we use local, staging, and prod environments) for your Serverless function, also set fax_gateway.staging.twilio_phone_number and fax_gateway.local.twilio_phone_number so you use a different fax number for each environment.

Check That You’re Registered, And Get A Mail-In Ballot!

You don’t need to be an expert in sending faxes to register to vote and request your ballot. You can use VoteAmerica to double-check that you’re registered to vote, and to request a mail-in ballot.

Building a Reliable Fax Pipeline with Twilio and AWS to Expand Access to the Ballot Box was originally published in VoteAmerica on Medium, where people are continuing the conversation by highlighting and responding to this story.

On October 20, 2020, I wrote to you to say that we needed to close a $985,000 funding gap. As of an hour ago, that gap is down to $19,000. Closing this gap in 9 days wasn’t easy. Justin and I made more phone calls than any two humans should ever make in a week, the texting team recruited a mess of volunteers to reduce labor costs, and the program/tech/comms/outreach teams raised $258,000 online. If you chipped in over the past 9 days, thank you. And if you haven’t chipped in yet, don’t worry: now that we’ve closed one gap, I’ve decided to open another. With 5 days and 5 hours left to go, VoteAmerica is launching another campaign tomorrow morning targeting unregistered voters. Here’s why:

34% of unregistered voters can cast ballots on Election Day. 34% of unregistered voters can show up with ID, fill out a voter registration form, and then vote on November 3, 2020. This is more than enough potential new voters to make up for any voter registration shortfalls we’ve seen so far in 2020, and it’s more than enough to ensure a decisive outcome to this presidential election. It’s called Election Day Registration (EDR) and 21 states offer it.

If you’ve never heard of EDR, you’re not alone. Most Americans have no idea that EDR exists. This is a marketing issue: election officials lack communication budgets, and partisan groups focus all of their efforts on registered voters. Also EDR is relatively new, and the American political class isn’t exactly known for innovation. So what we have here is a breakdown in communication, and a handful of days in which to fix that. Fortunately, VoteAmerica is a rapid response organization, and there’s nothing the team enjoys more than adding new programs to the docket.

Here’s an Election Day Registration primer:

• 21 states have EDR: Alaska, California, Colorado, Connecticut, District of Columbia, Hawaii, Idaho, Illinois, Iowa, Maine, Maryland, Michigan, Minnesota, Montana, Nevada, New Hampshire, Utah, Vermont, Washington, Wisconsin, Wyoming.
• For all intents and purposes, these states don’t have registration deadlines.
• There are 40 million unregistered voters as of yesterday. 13.4 million (34%) of unregistered voters live in EDR states.
• The EDR process is straightforward. Citizens go to a designated location (generally their polling place), show their ID, fill out a voter registration form, and cast a ballot.
• Citizens in EDR states have very few excuses for not casting ballots: if they can get to the polls, they can vote.
• GOTV drives in these states are easier, because you can target any citizen, not just registered voters.
• Here’s a spreadsheet of EDR data that shows which states have EDR and how many unregistered voters live in each state . (It’s also got some VoteAmerica specific information, but it’s not confidential.)
• EDR campaigns present a chance to expand the electorate at a reasonable cost. The cost per net new vote in EDR states is potentially lower, because organizations don’t need to run separate voter registration and voter turnout drives. (Post election we can talk about why progressive funders are spending millions of dollars running voter registration drives in these states in the summer and fall when literally any citizen can go vote on Election Day, but I’ll save that conversation for another email.)

We’ve been laying the groundwork for this campaign over the past few weeks, and are doing our best to maximize the resources we already have on hand. Here’s what we’re up to:

• We’ve integrated EDR information in all of our tools and our website. Most groups will want to bookmark our comprehensive EDR page. It lives here: www.voteamerica.com/edr.
• We’ve integrated EDR messaging into our campus program. Every ad targeting a student in an EDR state practically shouts at them about EDR.
• We’re sending 2 waves of emails to our 3 million person list, letting everyone know which states offer EDR and asking them to pass along the information. We’re not limiting ourselves to EDR states because this election is national, and everyone knows someone who lives in an EDR state.
• We’re sending 2 SMS messages to our 1.7 million person SMS list, also informing them that EDR is an option and asking them to pass along the message.
• Our comms team is pitching like mad to local media and national media, and I’m taking every TV, radio, and print opportunity that comes my way. Earned media is the best media, and reporters and journalists are really excited about EDR (real talk; everyone loves EDR).

We’re also adding a new program on top of our existing efforts, and this program is relatively inexpensive.

• While 21 states have EDR, VoteAmerica is focusing on 6 states that have competitive elections: Colorado, Iowa, Maine, Michigan, Montana, and Wisconsin. There are 2,970,000 unregistered voters in these 6 states.
• We are laying 2 more waves of P2P SMS messaging on top of our existing P2P messaging plan in these states, and we’re targeting all young people, people of color, and unmarried women, regardless of current registration status. We’re adding new waves of messaging instead of cramming this new information into existing messages because we don’t want to overwhelm voters (studies show that providing too much information at once reduces turnout).
• The program will cost $168,500 and will reach almost 3 million potential voters. We’d be stupid to not run this program.

This program will work.

We’re too frugal to spend money on things that are unlikely to move needless, and we actually attempted this on a small scale once before. In 2018, I ran a small EDR P2P SMS program with Professor Chris Mann. This program increased turnout by 1.35 percentage points, and yielded net new votes at a cost of $45 per vote (anything under $100 is considered very cost efficient). This positive effect resulted from sending only one wave of messaging. We’re sending two this time around.

Isn’t this election already in the bag?

No, absolutely not. As voters flock to the polls, conservatives flock to the courts in an attempt to keep election officials from counting ballots. As I told a reporter earlier this week, there are three ways to win an American election: you can win based on votes (democracy), you can suppress the other side’s votes (suppression), or you can refuse to count the ballots (rigging). Since vote suppression tactics are no longer working, conservative groups are leaning hard into option 3. I’ve lost track of the number of lawsuits designed to curtain ballot counting, but the general push is to not count mail-in ballots that are postmarked by Election Day but arrive afterward. A few issues with this. First, voters are being punished for USPS delays, which are out of the voter’s control. Second, USPS delays are due to a partisan attack on USPS service, designed specifically to undermine mail-in voting. Third, voting has already started, which makes changing the rules now absurd: we wouldn’t tolerate changing the rules of a football game during the third quarter of the Super Bowl, and we shouldn’t tolerate this during a presidential election.

VoteAmerica isn’t going to stand for this nonsense. If bad actors are trying to reject mail-in ballots, we’ll increase the number of in-person ballots. EDR lets us do this at a reasonable cost and with modest sums of money and time.

If you have any money left to give, this is what you should fund. If you’re anxious about the election, this is what you should fund. If you’re exhausted from the past 4 years, this is what you should fund. We added this program at the 11th hour, because we have faith that you will come through with the capital, while we deliver on the program.

• You can donate online here: https://secure.actblue.com/donate/voteamerica-medium
• If you plan to give more than $1000, this document has details on wire transfers, DAFs, and stock donations.
• If you want to give, but you need a few days to get the money together, just email me. Our invoices won’t be due until post election, so if you commit in writing to moving money by November 9, we’re good.

Clear eyes, full hearts, put on your mask, and let’s do this.
Debra

P.S. There are some definite upsides to this clusterfuck of an election: all of the work that went into ensuring long lines at the polls isn’t paying off. Folks are bringing folding chairs and snacks and audiobooks to listen to, and then patiently waiting. In any other year this nonsense would deter voters, but not this year. Folks are happily posting online about how many hours they waited to vote. This is a badge of honor in 2020. It warms my heart, and it sends a clear message: “we know you’re trying to keep us from voting, and we won’t let you.”

34% of Unregistered Voters Can Cast Ballots on Nov 3, 2020 was originally published in VoteAmerica on Medium, where people are continuing the conversation by highlighting and responding to this story.

About once a week I send an email to the VoteAmerica Insiders list. At the request of many of the recipients, I’ve decided to cross post these emails on Medium. Feel free to share and to donate: https://secure.actblue.com/donate/voteamerica-medium

VoteAmerica is a voter turnout organization, not a voter registration organization. I say this because over 85% of Americans are already registered to vote, so simply signing up new voters won’t solve the voter turnout problem. Americans don’t vote because it is objectively harder to cast a ballot in the United States than in any other nation with democratically elected leadership. So the key to increasing turnout is identifying and removing roadblocks to participation. Sometimes the roadblocks are knowledge-based, meaning that citizens don’t know when and where to vote. Oftentimes, the roadblocks are created by elected officials who proactively seek to suppress turnout. VoteAmerica tackles both issues: we provide rigorously fact-checked information to voters and best-in-class technology while making it abundantly clear to elected officials and government employees we will not tolerate voter suppression activities. Democracy is not a spectator support, and we are not fucking around.

Messing with Texas

For the second time in as many years, the team has stepped up to build critical election infrastructure for the State of Texas. In 2018, we built an “online voter turnout” tool for Texas that created a paperless workflow for Texas voters by submitting completed voter registration forms via fax and then printing and mailing a copy to the County Clerk within 4 days. This was a smart, inexpensive solution to a truly stupid problem, namely Texas’s refusal to build an online voter registration system for voters. We were quite pleased with ourselves for this one. The Texas Secretary of State (SOS) was less pleased, and almost immediately accused us of breaking the law by following the Texas law as written. If you’ve been on my update list for a while, you know how this played out. The Texas SOS made some vaguely threatening statements, we shut down our tool, and Perkins Coie immediately sued. The lawsuit is still active, and eventually, we’ll win and Texas will have online voter registration (which is the case in 41 other states and DC). In the meantime, we’re finding interesting ways to mess with Texas’s commitment to voter suppression:

• First, we added support for printing and mailing voter registration forms for people who used our register tool and didn’t have a printer. We didn’t have the funds to turn this on for all 50 states, but we did turn it on for Texas.
• Second, we sent 2.5 million P2P SMS messages to unregistered Texas voters encouraging them to vote. Ultimately, 13,754 Texans registered using our tool and asked us to print and mail the forms for them.
• Third, we sent paper voter registration forms (along with a pre-stamped and addressed return envelope) to 39,074 people who moved to or within Texas. We focused specifically on renters because renters are significantly more likely to be members of the Rising American Electorate (young voters, POC, and unmarried women) and are also significantly less likely to vote.
• Finally, we built a ballot tracking tool for Texas. The tool lives at TXballot.org, and went from conception to launch in about 72 hours. While the website says VoteAmerica, it was actually a joint project of a handful of humans, all of whom share a commitment to increasing voter turnout. The Texas SOS publishes this data, so they’re technically on the team, but I fully expect them to threaten to sue us any day now for using public data to create a public service. In the meantime, the Texas Democrats are overjoyed and are promoting the heck out of this tool.

Speaking of lawsuits

• The Florida online voter registration system crashed on October 5, 2020, which was one day before the voter registration deadline. Within 12 hours, VoteAmerica retained the services of the Campaign Legal Center and prepared to file a state (or Federal) lawsuit as a way of forcing Florida to extend the deadline. In this instance, the threat of a lawsuit was sufficient, and Florida extended the deadline.
• As of about five hours ago, it looks like we’ll be filing a preliminary injunction against the state of Pennsylvania for attempting to purge the voter file of individuals that the Social Security Administration has deemed “deceased.” The SSA, bless their hearts, has a long and well-documented history of declaring people deceased who are very much alive and well.
• There’s also a chance that we will file a second lawsuit against Pennsylvania. Their online voter registration tool requires citizens to enter their PA driver’s license number, but this system is also rejecting valid PA driver’s license numbers.
• More on both of these lawsuits as they unfold.

Our GOTV programs are hot like fire.

• WHERE TO VOTE tool. In partnership with USDR (the U.S. Digital Response), we built a WHERE TO VOTE tool that locates early voting locations, official dropbox locations, and election day polling place locations for voters. The tool pulls from multiple data sources, including the Google Civic API, and is the most comprehensive tool available. It’s also embeddable, so feel free to embed it on your sites. Polling places move, early voting locations close, and nefarious actors put up fake ballot drop boxes, so having a trusted source of this information will increase turnout. You can send traffic to our site (www.voteamerica.com/where-to-vote) or you can embed it on your website. The code is here: https://docs.voteamerica.com/embed/free/
• P2P SMS: The VoteAmerica P2P team is shattering all known P2P records. The team has sent over 75 million (not a typo) messages to low and mid-propensity voters, and is currently sending over 3 million messages a day with the ability to scale up to 6 million without issue. They spent the first few weeks focused on voter registration, and are now all-in on early voting reminders and vote-by-mail chase. The program has targeted 25 million voters in 20 different states and 206 different congressional districts.
• Text-based voter helpline. Voters who use the VoteAmerica tools are automatically opted into our email and SMS-based election reminders program. Within days of turning this program on, the team noticed that dozens of voters were replying to our shortcode (48298) with requests for help. So along with volunteers from Google and Slack, the team built a text-based voter helpline. Anyone who texts HELPLINE to 48298 is connected with a trained volunteer who provides help via text. This is already cool, but the really cool part is all of this runs through a Slack Workspace. Inbound SMS requests for help are routed to state-specific Slack channels. Trained volunteers are assigned to specific states, and reply to the voter questions from within Slack. The Slack workspace has a Twilio integration which converts the Slack response into an outbound SMS message. The volunteers receive and send messages from within Slack. The entire thing is very, very cool, and also effective: in the 7 days since launch, over 100 volunteers have been trained and 1,000 voters have been helped to date. We expect the requests for help to soar as we get closer to the deadline. (If you’re interested in learning more about the software, check out www.voterhelpline.org. If you’re interested in volunteering, sign up at www.voteamerica.com/volunteer. )
• Campus outreach. Despite COVID-19, the overwhelming majority of 4-year schools are providing an on-campus experience for students this year. Dorms and campuses are open, even if classes are held via Zoom. The VoteAmerica team has been running campus programs for years now and has consistently seen a turnout boost of between 2 and 6 percentage points, based on the length and breadth of the program. We’re currently active on 227 of the most competitive college campuses in America, in a comprehensive program targeting 2,710,614 students. That is a mind-boggling number of students.
• HBCU outreach. Our campus program includes a historic investment in HBCUs (historically black colleges and universities). We are running programs on 94 out of 107 HBCUs. This is the largest coordinated HBCU program in history. Whenever possible, we bought out all of the campus media inventory. And when those campuses didn’t have inventory, we worked with Flytedesk to create and sponsor inventory from scratch in the form of campus voter guides.
• Billboards. As noted in a past email, 37% of Americans surveyed in October 2016 didn’t know when election day was, which helps explain why 41% of Americans didn’t vote. Billboards and other out of home (OOH) media are an inexpensive way to fill this educational gap. We’ve put up almost 3000 billboards over the past 3 years, and have shown that billboards provide a consistent and cost-effective boost to turnout. This year we’re focusing on digital billboards (lower production costs) in three critical states: Arizona, Georgia, and North Carolina. This billboard promotes both early voting and Election Day voting.
• And GOTV campaigns are working. Americans are voting in record numbers in 2020, and they are undeterred by voter suppression tactics such as limiting drop boxes to one-per county (fuck you, Texas), or 10-hour lines at early voting centers (fuck you again, Texas, and Georgia and North Carolina as well).

Funding opportunities.

• VoteAmerica team is still actively raising and deploying capital in a cost-effective fashion. As noted above, while many teams are scaling down, we’re just getting started. Americans are registered to vote, and now we need to pull them across the line. And we need to do this while facing record election chaos. There is a virulent stream of disinformation coming directly from the White House; a deadly airborne virus that has crippled the economy and killed over 220,000 Americans; hostile foreign interference from Russia; nefarious actors who are doing everything in their power to keep votes from being counted (from lawsuits designed to stop ballots from being counted, to installing fake ballot drop boxes, to setting legitimate drop boxes on fire).
• VoteAmerica has a $796,000 funding gap that we need to close this week in order to run our P2P program for the 7 days leading up to Election Day. We need to raise another $189,460 to run our billboard campaigns in the three most competitive states in the nation for the 7 days leading up to Election Day. In other words: we do not have the funds we need to run the most cost-effective voter turnout programs in America during the most critical voting period in American history.
• This is unfortunately not uncommon in the nonpartisan voter turnout space. Although the lion-share of voter registration and turnout work is done by nonpartisan organizations, the overwhelming majority of funding goes to partisan groups and their media buyers.
• At the same time, we have the opportunity to expand our current programs. We can easily expand our billboard program into more competitive states and congressional districts. And we can go bigger on our campus program. For $1.2m we can add 4 touch-points to 1.5 million students in Arizona, Wisconsin, Michigan, Pennsylvania, Florida, North Carolina, Georgia, and Texas. For another $600k we can text 640,000 students right before Election Day. We have a partnership where we can access student directory data — so we can text entire student bodies in states with Election Day Registration. These are unregistered voters that no one else is talking to.
• Here’s a deck on the campus program expansion: https://docs.google.com/presentation/d/1rnW2ZEcxCMGIwQZ_P0mB98cxY-JgmQ1coqYBAfIoqiw/edit#slide=id.ga3fe1339aa_0_49

My grandfather used to say that if you have a problem that can be solved with money, and you have money, then you don’t have any problems. Here we have two problems that can be solved with modest amounts of money, and a chance to invest additional money in two of the most cost-effective voter turnout programs ever run. And every dollar donated to VoteAmerica is tax-deductible.

• You can give online at this link: https://secure.actblue.com/donate/voteamerica-medium
• If you plan to give more than $1000, this document has details on wire transfers, DAFs, and stock donations: https://docs.google.com/document/d/1dpv0GLPRYPjA7zCw-zNyQU8-PdrorsfIV31Bh6AREI0/

14 days left to go!
Debra

14 days to go and things are on fire! was originally published in VoteAmerica on Medium, where people are continuing the conversation by highlighting and responding to this story.

Sent: October 5, 2020

Hey friends.

An October 2016 study conducted by the American Council of Trustees and Alumni (ACTA) revealed a troubling lack of knowledge about basic aspects of elections and governance in the United States. The most astonishing statistic was this one: 37% of Americans overall and 22% of college graduates could not correctly identify when Election Day is held.

I suppose I should be alarmed by this statistic, but I’m actually energized by this. 37% of Americans didn’t know when the election was in 2016, and 41% didn’t vote. If citizens aren’t voting because they don’t know when to vote, we can increase turnout by telling Americans when and where to vote. This may seem like pure conjecture until you remember that this is exactly what my team and I did during the Alabama special election of 2017.

If you missed that one, let me catch you up. The fine folks in Alabama held a special election in December 2017 to replace Jeff Sessions. The only reason to hold an election during an off month of an off year is to purposefully reduce turnout. As noted above, Americans barely participate in regularly scheduled presidential elections. It goes to follow that turnout would plummet during a one-off election held in December of an odd numbered year. Which is why experts predicted low turnout. Experts further predicted especially low Black voter turnout, which actually made sense since both candidates were running programs targeting white voters.

A few weeks before the Alabama election, a funder called me and asked if I thought it was possible for a Democrat to win this race. I assured him it was mathematically possible since 32% of the registered voters in Alabama are black, and these voters support the Democratic candidate 98% of the time. Even the white voters down in Alabama vote Democrat18–23% of the time. So, if one were to increase Black voter turnout while allowing white voter attrition, then yes, it was possible that the Democrat could win this race. It would be close, but it was possible. The donor told me that his “political guy” said that this would cost $20 million dollars. I responded that $20 million meant broadcast TV, and that there was no way in hell that anyone was flipping Alabama using broadcast TV (Fun Fact: 51% of Americans under 50 haven’t watched a single minute of broadcast TV in over a year, so realistically, broadcast TV is an expensive way to reach people who are already going to vote).

The donor asked what I would do in this case. I said that I would stay far away from broadcast TV, and that I would lean into low cost inventory that would allow us to reach every registered Black voter multiple times with a reminder that there was an important senate race happening on December 12, 2017.

By now most of you know how this all turned out. Two individual donors chipped in just under $800,000, and we used this money to run the largest GOTV campaign in Alabama history. We bought every single billboard and transit ad in a majority Black county; we sent multiple waves of peer-to-peer SMS messages; we ran three weeks of radio ads on Pandora and Spotify; and we sent a wave of direct mail to every household with at least one registered Black voter. The message was simple: we simply advertised the date of the election and let folks know that we were expecting record-high turnout.

Record high turnout among the Black voters that we targeted is exactly what this campaign produced. Black voters turned out at levels on par with 2016 levels, white voters stayed home, and a Democrat was swept into office on a sea of Black votes. This is the exact opposite of what DC experts predicted, but it’s very much in line with what any marketing expert would have predicted. If you market voting like a product, and you are relentless about it, you can overcome resistance to voting. Especially if that resistance is caused by simply not knowing when Election Day is.

I’m going someplace with this, promise.

Americans don’t vote because voting is harder in the United States than in any other nation with democratically elected leadership. All 50 states have different rules for participating. The date of the election changes every year. We vote on a workday for no apparent reason whatsoever. Polling places move, often at the last minute. State election websites crash on key voting dates (Pennsylvania and Florida’s online voter registration websites have both crashed during the past 24 hours). Voting laws change repeatedly, sometimes multiple times in the same week. States demand voter ID, and then reject perfectly valid IDs as a way of suppressing votes. Polls close while folks are still in line. Folks are purged from the rolls for being dead even if they are very much alive. Etc.

These are all very real and very complex problems to solve, but it turns out that Americans also don’t vote because they don’t know when to vote. And this is why I am so cheerful as I type this email. Because it is incredibly easy — and cheap — to tell Americans when to vote. (Yes, I’m talking about billboards).

Over the past few years, I have put up just under 3,000 billboards informing people of when to vote. And I don’t just put up a few billboards: I buy every single billboard (and transit ad) in an area and fully saturate the community with a VOTE message that includes the date of the election. Before I purchased a single billboard, I did extensive research on the efficacy of OOH (out of home) advertising as a marketing channel, so I already knew that OOH worked and that the CPM (cost per 1000 views) was cost effective when compared to any other marketing channel. Now I know that OOH also works to produce new votes in a cost effective way. I know this because I’ve been working with Professor Chris Mann and Professor Scott Minkoff to measure the efficacy of billboards as a GOTV medium. Professor Minkoff’s conservative estimate is that the billboards produce net new votes at a cost of $23 per vote, or 43 votes per $1000. 8 votes per $1000 is considered a solid return; 43 votes per $1000 is the electoral equivalent of a fire sale on votes.

The average billboard costs just $2,500 for two weeks, and our past research suggests that two weeks is all we need. Right now we’re raising money to buy billboards in census tracts that are at least 60% POC. We’re also raising money for other OOH (out of home) inventory, such as gas stations, supermarkets, and transit ads. The designs will be simple, as per usual, and the non-billboard ads will also include information about early voting and EDR (election day registration). You can check out some rough designs here: https://docs.google.com/presentation/d/1MvriyY_ZWhkcufY-6Y1i5lI6buRHQO9zLJLi8lgK1og/.

This non-billboard OOH costs $800 for two full weeks, which is so inexpensive that I would actually like to order all of it, tonight, before I go to bed. Putting it all together, here are some sample budgets:

• Arizona: 120 units, 36 million views, $240,000 dollars.
• Florida: 454 units, 129 million views, $1.5 million dollars
• Georgia: 791 units, 103 million views, $1.3 million dollars
• Michigan: 367 units, 72 million views, $215,000 dollars.
• North Carolina: 638 units, 49 million views, $860,000 dollars.
• Texas: 550 units, 185 million views, $2.2 million dollars.

But we are running out of time to buy inventory. The amount of available inventory has dropped by 50% in the past two weeks in Wisconsin, Pennsylvania, Ohio, and Michigan, and honestly, I’m more than a bit concerned about who is buying it (I haven’t been able to find out yet). The Bradleys, Einhorns, and Kochs like billboards as much as I do, and they specifically like to use them to suppress black and brown voter turnout.

Here’s what a Charles Koch billboard looks like:

Here’s what a VoteAmerica billboard looks like:

But again, I’m an optimist, and I am confident that we can snap up the inventory and use it to increase voter turnout, while also preventing nefarious actors from using this same exact inventory to suppress turnout.

If you’re ready to make a cost-effective investment in US democracy, that also happens to be tax-deductible, you should give to VoteAmerica. I really, truly, want to start buying the inventory right now because this is a very, very, very good idea.

• You can give online at this link: https://secure.actblue.com/donate/voteamerica-medium
• If you plan to give more than $1000, this document has details on wire transfers, DAFs, and stock donations: https://docs.google.com/document/d/1dpv0GLPRYPjA7zCw-zNyQU8-PdrorsfIV31Bh6AREI0/

Thank you,
Debra

37% of Americans don’t know when Election Day is. was originally published in VoteAmerica on Medium, where people are continuing the conversation by highlighting and responding to this story.

NB: Once a week or so I send an email to my personal contacts. People have asked me to cross-post on Medium so they can share. The emails are slightly edited (I remove my personal info), but otherwise unchanged.

Hey friend,

Well that debate was certainly energizing. I’m going to try to keep this email short so I can get back to work, since work is my outlet for stress, anxiety, etc.

Highlights of this email: the program is scaling rapidly, the team is at 29 people as of tomorrow, we are training volunteers twice a week, and we are still actively raising because our program scales with capital, not labor. And the team is filled with the sort of righteous rage (and optimism) that enables folks to work 7 days a week. Also if you don’t want to hear from me anymore, there’s an unsubscribe link in the footer.

Weekly Salon with Assembly Member Buffy Wicks and Mary Steenburgen.

• Date: Thursday, October 1, 2020
• Time: 4pm PT (7pm ET)
• VoteAmerica is excited to welcome Buffy Wicks, political badass who has been at the forefront of the discussion of maternity rights and COVID, and Academy Award-winning actress and songwriter Mary Steenburgen. We’ll chat about the upcoming election, women voters, and what it’s like serving in the CA assembly while raising two very young children.
• Buffy and Mary’s bios are on the RSVP page. It’s free to attend, although we won’t say no to a donation!
• RSVP here: https://secure.actblue.com/donate/voteamerica-2020-buffywicks
• You can share this event publicly! Buffy and Mary are so much fun!

Program highlights:

• Our P2P texting team will have sent over 30 million messages to voters in competitive districts by the time I hit send on this email. The team is now sending over 2 million messages a day. This is an astonishing number of messages. To date we’ve been heavily focused on voter registration and VBM recruitment. As of yesterday, we’re focused on early voting reminders and VBM chase. Earlier this year we had assumed we’d be leaning more into VBM recruitment than early voting, but that was before a partisan attack on the USPS crippled mail delivery. With mail delivery uncertain, it seems best to encourage voters to start casting ballots now. There’s also a chance that — if enough people vote early — we might actually know who the winner is on election night. Regardless, we’ll be leaning into early voting for a bit, and then we’ll move on to Election Day Registration (EDR), Election Day Voting, and post-election ballot curing.
• Speaking of the P2P program: our volunteer program has taken off. We’re running at least two trainings a week now, and volunteers have sent over half a million messages. If you need an outlet for your rage and/or anxiety, I encourage you to sign up. We can train your entire school or company at once, and put you all to work. Texting is fast, easy, and effective, and volunteer labor reduces our costs, which lets us run bigger programs. You can sign up here: www.voteamerica.com/volunteer
• Campus program: our campus program is running at 150 schools, which allows us to reach almost 2 million young voters who attend schools in the most competitive districts in America. By mid-October we expect to be live on at least 120 more, including all 94 HBCUs (Historically Black Colleges and Universities). Here’s a fun fact about the schools: 94% are providing a residential experience, even if the classes have gone online. So while learning may be virtual, the students are still physically on campus. Which is great for those of us in the voter turnout space.
• Tech highlights: our tools have now been used over 2 million times, including 500,000 times last week alone. On National Voter Registration Day, the tools were being used over 300 times a minute, without issue or downtime. Our tools are embedded in over 300 websites, including on CreditKarma.com. The tools are powering phenomenal email and SMS list growth: as of this morning, our opted-in email list was just north of 2 million people, and our opted in SMS list was 1.2 million. SMS reminders are a powerful way to increase turnout, and we are looking forward to hammering our list with reminders to get out there and vote. If you want to embed the tools on your sites, please sign up here: premium.voteamerica.com.
• Election Information API: our research team maintains our election information API. This means that these long-suffering souls need to keep abreast of election law changes. So far they’ve tracked 108 changes this year. If you need reliable election data, that is free for you to consume and use, learn more here: https://docs.voteamerica.com/api/
• Media highlights: I’m putting this under program because media drives money, money drives program, and program drives impact. (This took me years to learn because I am a slow learner). Despite being a very young organization, VoteAmerica has reached over 250 MILLION media impressions through a strategic and cost effective earned media campaign. It feels dorky to link to the clips, but we’ve been featured in a number of high profile outlets, including the New York Times, Fast Company, the Wall Street Journal, CNN, MSNBC, and a few dozen others.

We are still scaling programs and raising funds

• If you’re worried that it’s too late to invest, let me assure you that it’s not. We are still very much raising and deploying capital. And we’re doing it in a cost-conscious way, because we’re not interested in setting money on fire over here. Here are some things we’re raising for:
• The campus program. We book the media 2 or so weeks before we run it, and right now the program is funded through mid-October. We have a $300k gap that we’re trying to close. As I’ve noted in prior emails, we’ve been running campus engagement programs for a few years now, and the program consistently produces between a 2 and 6 percentage point increase in turnout. There are no guarantees in the voter turnout game, but I can tell you this: if someone had invested $1.5 million dollars in this sort of campus outreach program in 2016, a different person would be sitting in the White House today.
• Election Day Registration: 21 states offer in-person registration after the mail-in deadlines, including key states such as North Carolina, Michigan, and Wisconsin. You can read all about this here: www.voteamerica.com/election-day-registration. Most folks have no idea that EDR exists, which is where we come in. At a minimum, we will be sending millions of text messages about EDR, but we want to go bigger. We specifically want to invest in billboards, radio, postcards, and online advertising to spread awareness. EDR is the antidote to suppressive voter ID laws: studies show that EDR increases turnout even more than voter ID laws decrease turnout, and it’s in our collective interest to use EDR to drive even more voters to the polls. If you only fund one thing this year, fund our EDR work. Ignoring EDR is leaving votes on the table for no clear reason: these are our votes to pick up, we just need to do the work. The best way to advertise this is with billboards, and we are rapidly approaching a hard deadline with billboards. The drop dead date on almost all billboards is October 9, 2020. Don’t worry: I’ll send a separate email about this.
• Print and mail: We are now printing and mailing forms and pre-addressed, pre-stamped envelopes to voters who use our registration or VBM tools and who don’t have access to a printer. There is a limited window here, since we will turn this option off as we get closer to deadlines, but this option is very popular since home printer ownership is the low single digits.
• Recent movers: we are getting real-time mover data from a company that handles change of address information at least 28 days before it appears in the NCOA (National Change of Address) list. Rather than wait for these folks to realize they need to register to vote, we are proactively sending them registration forms that will arrive shortly after their move. We’re focusing specifically on renters who are moving into competitive states. Renters are a more progressive voting block than homeowners for reasons that are beyond the scope of this update. As the registration deadlines pass, we will switch to postcards to these folks if they live in EDR states, advising them that they can register at the polls. Given recent shenanigans with the NCOA data, this program is more important than ever. In case you missed it: the USPS failed to update addressees for at least 1.8 million people, which is a great way to undermine mail-based registration programs targeting movers. But since VoteAmerica is in the unique position of not relying on NCOA for our mover program, we can still run a program that we are highly confident in.

Folks keep asking me what the funding gap is. It’s less a funding gap and more of an opportunity gap: our programs scale with capital, and we can continue to invest and scale as right up until (and after) Election Day. Our GOTV program doesn’t stop until a full week after November 3, since we’ll use the post-election time to focus on ballot curing. We could stop raising today and still have run one of the country’s largest voter turnout efforts, but we honestly don’t think that’s enough. Whatever happens on Election Day, we want the margin of victory to be wider enough that the outcome is decisive. And that will only happen if VoteAmerica and other C3 orgs have the funding needed to run truly large scale programs.

Here’s how to give:

• You can give online at this link: https://secure.actblue.com/donate/voteamerica-medium
• If you plan to give more than $1000, this document has details on wire transfers, DAFs, and stock donations: https://docs.google.com/document/d/1dpv0GLPRYPjA7zCw-zNyQU8-PdrorsfIV31Bh6AREI0/

Let’s end this on a high note:

• 100% of donations that come in at this point go directly to our programs. Salaries, benefits, and what little overhead we have were covered via a 2019 grant. Every last dollar that you invest will go directly to voter registration, voter turnout, and voter protection. Donating will make you feel better about life, promise.

In solidarity.

Debra

P.S. If you’ve had trouble trying to reach Justin, it’s because his son arrived 9 weeks early. Baby and mom are doing well, and our fundraising operation is being run out of a NICU in Virginia. The rest of us appreciate the baby’s punctuality, and are happy that he will be able to vote in the 2038 midterm election. But this does mean that we are obligated to give him a functioning democracy as a birthday present. So if you need another reason to give, I hope that Cortland Parcell Jones is that reason.

35 days and 2 presidential debates left to go. was originally published in VoteAmerica on Medium, where people are continuing the conversation by highlighting and responding to this story.