[sinak]

Hey there, I worked on this.

BSD, BlackBaud, and Salsa all deliver messages to congress, but only for advocacy organizations who are willing to pay. Beyond OpenCongress, I don't think there are any tools that make it easier for constituents to write their own non-cookie-cutter messages to Congress. Change.org, for example, doesn't deliver emails.

The reason why Congress receives millions of messages is because advocacy organizations send millions of form letters. Congressional staff already have plenty of tools for separating those form letters from real, constituent-written letters. I can dig up their names, but they're built by high-level contractors.

Finally, one of our plans for Democracy.io is to measure response rates from representatives, and to use that to release a public report on how well MoC respond to real constituent messages. Both in terms of timeliness, and the relevance of their response.

[cjoh]

> Congressional staff already have plenty of tools for separating those form letters from real, constituent-written letters. I can dig up their names, but they're built by high-level contractors.

Intranet Quorum. Last updated in about 1996. It's probably what 95% of staff uses, and it's woefully inadequate. Clearly you talked to zero congressional staffers in the creation of this software. Which is a shame. Because if you had, your software would be effective.

Heck, even if you'd read one of the dozens of reports from the Congressional Management Foundation _FROM 10 YEARS AGO_ http://www.congressfoundation.org/index.php?option=com_conte... you could spot the problem here.

Whether advocacy organizations "send" the emails or whether they're sent by an individual has no bearing on either the legitimacy of the message, or whether or not they will remain unread.

I'd encourage you to hop on a plane, and go visit some congressional staffers and ask them how they'd like to receive your messages. Then build a tool starting from there. Like it or not, they're the customer.

[sinak]

Clay, I have read those Congress Foundation reports. And we regularly visit congressional staffers in DC. I'll also try not to resort to all caps in responding :).

Obviously your product, ScreenDoor, addresses some of the concerns you're raising around better data collection. Glad to hear you're so enthusiastic about that approach.

The EFF isn't planning on selling software to members of Congress. However, we are considering offering users the ability to make their comments public, which'd create a large public dataset about messages sent to Congress. If we decide to go ahead with that, the open dataset would then be available for anyone to run analytics on and create reports for members of Congress.

Improving the ways people contact Congress is one piece of the puzzle. The other part is making sure that people get better responses, and we hope to work on that too.

[cjoh]

> Obviously your product, ScreenDoor, addresses some of the concerns you're raising around better data collection. Glad to hear you're so enthusiastic about that approach.

Not sure what that's about other than it being a strange comment to make.

> The EFF isn't planning on selling software to members of Congress.

Who said anything about it being commercial? I think that if you talk to a member of Congress, or an LA or SA inside of the hill, they'll all tell you the same thing: we take electronic messages and basically throw them away. That's nearly universal feedback. And before we start thinking it's out of cynicism or dislike of the constituent, it isn't out of malice. It's because if they started reading every message they receive right now, they'd be stuck in July of 2015 for the next 10-15 years with the resources they have.

So my point is: why bother with increasing their volume. It's a bit like shipping gasoline to a forest fire.

The reason this is a problem is because it drives up apathy and cynicism. People will send more messages, get worse responses, and continue to be validated with the idea that Congress doesn't care about what they have to say.

[sinak]

> So my point is: why bother with increasing their volume. It's a bit like shipping gasoline to a forest fire.

Isn't this exactly what you did at BSD? Enabled advocacy organizations to send millions of form-letter to Congress, making real constituent contact get lost in the flood?

I've already addressed how we plan on incentivizing members of Congress to write better responses, but again: we're looking at an open dataset of comments, and reports on timeliness and content of responses.

[cjoh]

>Isn't this exactly what you did at BSD? Enabled advocacy organizations to send millions of form-letter to Congress, making real constituent contact get lost in the flood?

Yes. More than a decade ago, I admittedly had a hand in creating this problem, and now I'd love to see it fixed.

[sroerick]

>However, we are considering offering users the ability to make their comments public, which'd create a large public dataset about messages sent to Congress. If we decide to go ahead with that, the open dataset would then be available for anyone to run analytics on and create reports for members of Congress.

This was my first thought on seeing this project. That would be a great feature and could easily change the dynamics of "open letters" and petitions.

[konklone]

Respectfully, @cjoh, that's pretty out of touch. Blue State, Salsa, all of those vendors charge money and create necessarily limited ecosystems.

EFF and Sunlight collaborated to create the data ecosystem necessary for this project in order to change that dynamic and make it possible to mechanize the sending of letters to Congress for free. That EFF and Sunlight could justify spending a pretty amazing upfront investment of staff labor and resources to create that ecosystem demonstrates that the existing vendor-driven market was limiting who could participate.

> I'd encourage you to hop on a plane, and go visit some congressional staffers and ask them how they'd like to receive your messages. Then build a tool starting from there. Like it or not, they're the customer.

They're not the customer, and to treat them that way is a great way to ensure that Congress is never made uncomfortable. It doesn't matter how many staff complain to the CMF about it.

Congressional staffers definitely are the customer to Intranet Quorum and any attempts to compete with it. Those are the pieces of software that need to bend over backwards to take their needs into consideration.

For sending messages to Congress, the burden is appropriately on Congress to find a way to handle receiving them.

Responding to frustrated Congressional staff by reducing the information below between Congress and constituents would be patrician and elitist. Responding to frustrated Congressional staff by limiting the intermediaries to a narrow range of mostly partisan for-profit vendors would represent an interesting and dangerous form of capture.

[cjoh]

> That EFF and Sunlight could justify spending a pretty amazing upfront investment of staff labor and resources to create that ecosystem demonstrates that the existing vendor-driven market was limiting who could participate.

Not sure how Sunlight and EFF spending staff time on something validates or dis-validates a market.

> They're not the customer, and to treat them that way is a great way to ensure that Congress is never made uncomfortable. It doesn't matter how many staff complain to the CMF about it.

Let's be real: An 85% incumbency rate takes care of that. The reality is: Nobody is going to lose their office because they didn't reply to messages from Democracy.io. Moreover, no congressional _staffer_ is ever going to lose their job taking these messages and throwing them in the trash. There's not going to be some sweeping revolution in Congress because Congress wasn't representative enough.

The fact is that Congress' approval rating is 15% and their incumbency rate is greater than 85%. People already believe that their Congress doesn't listen to them, they also really don't like Congress and, it seems to me, are pretty much fine with that situation, at least electorally. The "This will make Congress uncomfortable" idea is one that's stale and old.

You can call this elitist or patriarchal. But my suggestion is that the EFF and Sunlight do exactly what you are doing at 18F (which I suggested to you 2 years before you took the job, scoffed at, and as I recall were quite argumentative about it. So...): go on the inside and work on problems from there, because if you're connecting firehoses to drinking straws, you're doing more harm than good.

Finally if you want to make it so Congress is well represented by its constituents, make it as easy and delightful to hear from your constituents as it is from a lobbyist. Until then, the lobbyists will always win.

[konklone]

> Not sure how Sunlight and EFF spending staff time on something validates or dis-validates a market.

They were previously customers of that market. They found it suited their needs better to replace the thing they were purchasing from.

> You can call this elitist or patriarchal. But my suggestion is that the EFF and Sunlight do exactly what you are doing at 18F (which I suggested to you 2 years before you took the job, scoffed at, and as I recall were quite argumentative about it. So...): go on the inside and work on problems from there...

I didn't scoff at working on the inside. I scoffed at the notion that working on the inside is required in order to have a valid opinion about what government should do. And that's something I continue to scoff at.

> because if you're connecting firehoses to drinking straws, you're doing more harm than good.

It's not zero sum. I'm glad for the companies that tackle the inside game here too. It's not the only game worth playing, and if democracy.io increases the pressure further for more people to work on the inside sector, then everyone wins.

> Finally if you want to make it so Congress is well represented by its constituents, make it as easy and delightful to hear from your constituents as it is from a lobbyist. Until then, the lobbyists will always win.

That's a pretty fluffy notion. Lobbyists can show up in person, and can get special access to events and meetings that ordinary constituents can't.

Sometimes that's earned by virtue of subject matter expertise, and a lot of the time it's just people who know each other. You can't make it more delightful to hear from a constituent than it is to get a call from a friend.

You can certainly argue democracy.io doesn't help with that particular problem, and that's fine. Arguing that it actually hurts democracy to help people send emails to Congress without having to go through a for-profit vendor is what gets into the territory of elitism.

[davidmooreppf]

Great thoughts & articulations, @konklone, wanted to add my support. David here, our non-profit PPF created Contact-Congress project on OpenCongress in 2010 and launched the first version publicly in 2011 to automate delivery of digital messages to Congressional webforms. Really excited to see the evolution of the Contact-Congress toolset, now with the launch of Sina & EFF's & team's Democracy.io.

Our primary goal was a positive user experience for OpenCongress visitors, who mostly arrived via search for official information about bills in the news or legislative issues they cared about - to walk these interested bystanders up the ladder of engagement, enabling them to write an informed letter (using info aggregated uniquely by OpenCongress) directly on the page where they learned about a bill in context (both its official status and social comments / plain-language summaries).

A secondary goal of Contact-Congress was to give users an immediately-shareable permalink to their letter to their members of Congress (if set to public, as opposed to private) - to demonstrate the potential of Congress treating constituent communications in an open CRM, a public queue, with tools to enable constituents in-district to share helpful resources, timely updates, and organize around their initiatives. My PPF Blog post on this potential for in-district organizing, from May 2014: http://goo.gl/gCS0fy

A tertiary goal of Contact-Congress was to demonstrate public demand for open letters and open priority lists for their Congressional offices, and highlight the potential of open data standards for more constituent communications. This is what I've described as the greatest #opengov potential in my nine years in this field (http://goo.gl/crVivS) - better listening tools, as per @cjoh's priority request - I wrote: "Government staff have new access to dynamic, data-rich dashboards for opinions and feelings of their constituents – crucially, that integrate with their official CRM & CMS solutions, so a bigger-picture is generated, gathering more public feedback & specific expertise." Exactly such an open data standard already exists and has been successfully tested, a few years ago, with a U.S. Senate office. So what's holding back development and adoption of such a standard for delivery & verification of more communications, such as petitions, questions, volunteer offers, public testimony, community events? In my view, it's simply a lack of charitable funding support for open-data infrastructure for engaging with government offices - I believe this can implement more & better constituent messages to Congress, including geolocation features for constituent verification in cases where implemented, and better analytics tools for Cong. offices to respond to messages without being buried in a queue.

That’s a huge opportunity for non-profits and for-profit civic startups to provide the next generation of CRMs to government – investing in the infrastructure to make that possible now, by supporting the work of PPF and other non-profits, will help open up that data as widely as possible and ensure the participation tools developed on top of it are, at least in part, open-source for remixing. (My Dec. '14 blog post: http://goo.gl/AprZns).

[joelcollinsdc]

> I'd encourage you to hop on a plane, and go visit some congressional staffers and ask them how they'd like to receive your messages. Then build a tool starting from there. Like it or not, they're the customer.

This is the relevant contract vehicle... http://www.house.gov/content/cao/procurement/PDFSolicitation...

> Intranet Quorum. Last updated in about 1996. It's probably what 95% of staff uses,...

This is not accurate, at least anymore

[Balgair]

I mean, how can that be? How stupid easy is it to have a web app that sends out an email? Like every 2 person start-up out there has something like this for messaging them. Honest question here, are you really the first, in the year 2015, to do this for congress in one 'clearinghouse' type site?

As I have your ear, some more questions then:

How about a direct calling option? I've used your calling to eff then connect to a congresscritter function before and it works great. I also remember hearing that a phone call is much more important than some email.

What about accountability? My emails mean nothing if I still vote for them or have no other options. I signed up for GovTracker.us to send me email on my rep's voting, but I honestly cannot make heads or tails of the 300 odd bills they all voted for the last week. Some sort of editorial function is really needed if accountability is to be implemented.

[sinak]

You have no idea how complicated sending message to Members of Congress is.

There's no way to email legislators directly. The only way is to fill out their forms. Each form is unique, and many have CAPTCHAs.

To do that, we first had to develop an open source dataset that describes to a computer the steps necessary to submit messages via each contact form for 538 representatives. That was quite a lift, but was made possible by about 100 volunteers who chipped in. The dataset is called contact-congress, and is online here: https://github.com/unitedstates/contact-congress

Then we had to create a system to fill in the forms based on the dataset. The system we designed is called phantom-of-the-capitol and it does that by firing up 3 browser windows on an AWS instance each time a user submits a message. If the legislator has a CAPTCHA, we take a screenshot and return that to the user. The code for that system is here: https://github.com/EFForg/phantom-of-the-capitol

[Balgair]

WOW! That is terribly bad then! You are very correct, I thought it was just some email account they had in their name, not something like that. Great work arounds for the captchas too, nice idea. Honestly, I never even new they had forms that you had to fill, I typically call them on the phone and don't bother with email.

Still though, any thoughts on using the eff phone integration here to also prompt users to call the congresscritter?

Awesome work. Sometimes it is nice to see how much hard work goes into some simple idea.

[comrh]

Wow that contact-congress data set alone is awesome. Thanks for doing that leg work!

[kajecounterhack]

Wow thanks for explaining, that's a lot of good work.

[arenaninja]

Ahh yes, unitedstates/contact-congress. I remember when I tried to contribute, several reps had multiple topics listed in a <select>, but the <option>'s value was the same for multiple topics and the tests kept failing.... good times