So you want to reform democracy

Joshua Tauberer
Civic Tech Thoughts from JoshData
12 min readNov 23, 2015


[Note: This post was translated into French by Thomas Champion.]

Hello. You probably just wrote me an email that went something like this:

If a useful tool was put out there that was simple and engaging and closed the loop on opinion of the people and action by our representatives, it would get good traction.


The app will allow anyone with a smartphone, tablet, or the internet, to view bills moving through local, state, and national legislatures. We can vote on the bills after reading a summary. This allows 1) citizens to become aware and involved, and 2) give our representatives a very clear picture of their constituencies’ opinions.

These were recent emails. So you’re not the first to email me with an idea to:

  • Convince people that politics is interesting — because it really is.
  • Hold Congress accountable by measuring what the public wants.
  • Build a social network for politics/civics.
  • Have Americans draft legislation or comment on or vote directly on legislation in Congress.

And it’s great that you have your idea. I am thrilled you want to make our government better by rolling up your sleeves and doing something.

But your approach is all wrong.

You probably emailed me because you saw I’ve been working on this problem for a while. It’s true. Almost 15 years! So I know a few things.

(If you found your own way to this post — i.e. I didn’t send it to you after you cold emailed me — then welcome and I hope you get something out of it, but keep in mind this post wasn’t written for you so I’m not interested in why you think it doesn’t apply to your idea.)

If there was an idea that could ‘fix’ democracy, it would have been thought-up already.

Tens of thousands of people are working intensely on problems related to governance, like access to information (my area), voter registration, conflicts of interest, lobbying, organizing, polling, procurement reform, campaigning, campaign finance reform, criminal justice reform, investigations of corruption, gerrymandering, etc etc etc. Foundations and venture capitalists have spent hundreds of millions of dollars on making government and policy-making better, and many tens of millions of that just on use of technology.

Here’s you right now:

You’re not the first to think government could be better with tech.

If you wonder why you don’t see your idea already out there in the world, it’s because your idea didn’t work the hundreds of other times someone tried it. (I might have tried it: See my failures.)

Wasn’t GovTrack a good idea at the beginning?

When I started working on GovTrack, I thought I was building an accountability tool. If only the American public had more information they could head-off failures in government by voting more effectively in elections.

That was wrong. I was wrong.

Never to my knowledge has GovTrack affected an election. It’s never revealed a scandal, gotten anyone fired, or uncovered a failure in the legislative process.

And that’s totally fine because there are lots of other reasons why GovTrack is important, like engaging the public in understanding how Congress works, making the legislative process more real and accessible. [Update posted a year later: GovTrack has been a source for journalists and the impact certainly could have been to sway politics a little bit toward truth.]

On to your idea, which was probably:

Let’s tell representatives what their constituents think. Then they HAVE to vote with us!

Members of Congress don’t necessarily want to know what their constituents are thinking. And when they do, they have no trouble finding out:

They get a lot of feedback from constituents already. Many people write their representative and senators already, and each representative and senator’s office receives so many messages — hundreds on a slow day — that they can’t keep up. There’s no shortage of people telling Congress what they think!

Members of Congress also already do polling when they want to know what their constituents think. Whatever a Member of Congress wants to know, they can already find it out if they wanted to.

And no one wants to be polled — there’s nothing in it for people to want to participate. If your goal is to poll so many people that you’ll just know what Americans want, you’ll never get there: Only a small fraction of citizens will ever participate in a poll, and that causes selection biases. Unless you have a PhD in statistics and a background in polling, you’ll never overcome this better than the existing pollsters.

Everyone should vote on the issues before Congress — Let the people decide!

Congress takes more than 1,000 votes each year. Go through the actual list of votes and try to form an opinion on how you would vote on each. What do you think about the Wicker Motion to Instruct Conferees? Or the motion to approve the journal? Or the Defending Public Safety Employees’ Retirement Act? (Spoiler: That bill isn’t about public safety employees’ retirements!)

If that’s what you want people to do, try it yourself first and see how far you get.

The world is complicated and for good reason we elect representatives to make decisions so we don’t have to read 2,000-page bills. No only that, but the world REALLY IS complicated. What do you think the dividend rate should be for Federal Reserve member banks’ shares in their local reserve bank? This is an actual, serious issue in Congress that got a vote, and I bet you don’t know remotely what that sentence means. Now prepare for 999 more votes like that.

People are rationally ignorant. We want to live our lives. If everyone had to participate in every vote, there would be no time left to live. People don’t want a direct democracy.

We could delegate our vote to someone else.

You’ve just invented representative democracy! Well done.

We already do delegation, once every two years. First by delegating power to the political parties which draft candidates for office. Then we delegate to our elected representative to make governing decisions. And then our representatives often delegate their vote too — by taking instructions from party leaders on how to vote (rarely do they make their own decision)— and, finally, when a law is enacted (if approved by the President and not struck down by the Supreme Court — delegation upon delegation!), they delegate the details of policy implementation to government agencies which issue ‘regulations.’ There is so much delegation going on already.

How is your delegation system going to face the same challenges our existing delegation system faces? For instance: Selecting among delegates is hard. Did you vote in the primary election for your local water authority officer? How did you decide which candidate to pick? Start with the problems you see in our existing delegation system. Saying we simply need to delegate our vote misses how we’ve already done that, and that’s how we ended up where we are today.

Let’s make a social network for politics / civics.

Google has a social network called Google Plus. No one uses it. If Google couldn’t build a successful social network, neither can you.

Relatedly, people don’t wake up in the morning and say, “I sure hope I do something civic today!” Most people have other things to worry about. At most some might say, gosh it’s awful that [your favorite issue here] exists, maybe I can do something about [that]. Even you did that today. There are lots of civic volunteer opportunities you could have done today if you were just looking to be more civic, but instead you emailed me about … well, this is a generic response so I don’t know! But that.

We all have our pet issues. People aren’t looking to be civic, in a general way. They are looking to do something very particular that is important to them.

Ultimately you have to address an actual problem people are having in their lives. Spoons address the problem of conveying liquid from bowl to mouth without dripping too much. How is your idea like a spoon? What’s something some people want to do now that you can make less expensive? (You aren’t making anything possible with your idea. Everything is possible, with money. You only make things better compared to the next best alternative.)

Okay, so, then what?

Until you’ve worked 5–10 years in government or advocacy, you can’t see what needs change.

You can’t fix a broken machine until you study it and see how it works. Government isn’t something you can intuit. It doesn’t work like you were taught in school in 6th grade. Not even remotely.

Why is that?

People make sense of the world by learning principles: things fall down (not up), a smile means someone is happy, money is power, lard is unhealthy. We do this because the actual laws of the universe are a little too much for us to keep in mind, but by simplifying we end up getting things wrong sometimes — often without realizing we’ve gotten it wrong. A smile can indicate contempt but you might miss it if you aren’t paying close enough attention, right? And lard might actually be healthier than popular alternatives, like oils with transfats were in the ’90s.

We do this with our society too, with principles like: majority votes are fair, self-determination makes people happier, transparency reveals corruption. Unfortunately these are simplifications too, and, as with the ones above, they often lead us astray when we’re not careful: majority votes allow mob rule, self-determination doesn’t create equitable outcomes, and corruption moves under the shadows as soon as a spotlight is put on it.

These simplified ways of understanding governance and our civic lives are actually just really bad. There are more examples below of how bad they are. Cranking up one of those dials — adding more transparency or participatory governance — often has the oppose effect of what our intuition would tell us, or may negatively impact something else we care about, because the dials don’t exist in isolation. They are a part of an extremely complicated social dynamic of which you and I can only see a very small part of. Turn one dial and another one also turns that you won’t even be aware of until the next decade, but the result might be that everyone is worse off now. The simplifications are good ideals, certainly from a moral perspective, but they’re not good guidance for how to achieve better outcomes. Move past the simplifications by studying the system you want to change — study it over the course of years if not decades.

[The above section was added in June 2017.]

Where to start

If you want to be effective, start by getting a job working for someone else doing work you respect in this field. Then in a few years see if your original idea still holds up. You might decide that the way you can be most effective is working for the government! Alternatively, ask the people already working on these problems what they wish someone would work on — and do that.

If your idea is to change the way that laws are made, go become an expert in that. See the process first-hand by getting a job in the policy shop of a nonprofit. Don’t start working on your idea until you know where a lobbyists’s power comes from (it’s not money — if you think it’s money, you are not ready to work on this problem).

Individuals don’t hold power, groups do.

If democratizing policy-making means giving individuals the power to shape decisions, then it’s a meaningless idea. We’re a nation of about 300 million people. 0.0000003% of the population — one person — acting alone cannot shape policy, nor would we want that!

Power comes from organized groups. If you want people to have power, then you want to help them connect with others and teach them how to carry out effective advocacy together. That’s hard.

It’s not a technology problem. It’s not something that a slick website solves. Building power is a social, societal, institutional challenge.

Power is zero-sum. If you want one group to have more power, then you want to take power away from someone else. It’s two sides of the same coin. Are the people you want to take power away from going to sit idly by? How are they going to react to your idea? They have to be as much a part of your thought process and plan as the people you want to empower.

Be careful who you are empowering.

If you’re trying to use technology to help people build power, keep in mind that technology often excludes segments of our society with the least power and who are in need of the most help. Technology is something only available to those who can afford it (and the time to use it). If your target audience is people like you, your idea could easily exacerbate problems for people less well off than you.

Who are you trying to help? There are some real problems in this world. Minorities get killed for dressing the wrong way. The poor spend countless hours navigating government forms for assistance programs. Health care costs are constantly bankrupting people who are just unlucky. These Americans would love to be so privileged to have the time to debate national policy. Are you sure you’re trying to solve the right problem?

But- Don’t try to solve problems you don’t experience yourself without actually engaging with the people who have those problems. Every now and then some rich folks build apps about homelessness— don’t make that mistake, in whatever domain you’re working on.

There are a lot of real civic problems that need your help — consider one of those.

The most common mistake folks make is to think that there’s The Government that needs fixing, but there is no The Government. Our civic lives are driven by a network of tens of thousands of different systems that have evolved over decades and centuries to be what they are today.

Each of those parts of our civic lives — from elections to charities looking out for those in need — all need more people willing to help. Volunteer at a food bank or at a library to help your neighbors find the resources they need. Join a protest with people whose rights are being trampled.

Your “insight” into fixing government is something you invented because you had the luxury to do so. Meanwhile, poverty and discrimination go on.

Go to where the known problems are. Don’t make up new ones. And if you aren’t following the tradition of tzedakah — giving enough of your income to charity that it feels like a sacrifice (or whatever sacrifice makes sense for you) — then you need to ask yourself if you’re trying to make the world better, or just trying to feel good.

[The preceding section was added in December 2016.]

Become an expert first and find your unique contribution to the world.

The world is unbelievably complicated. You’re not going to make headway coming in as a generalist and hoping to fix all of government.

Everyone brings a unique perspective to this world. What is yours? It’s not your smart ideas. It’s your background, your passions, and your skills.

Apply your unique perspective to the problems you see. You should be proceeding with your life in a way that screams you. The problem you’re trying to address should scream you. Your manner of solving it should scream you. That’s going to mean honing your skills and focusing on a narrow problem with enough specificity to be meaningfully you. Why? Your experiences, particular passions, and skills are a resource — use them!

Use your uniqueness to become an expert. Become the go-to person for some obscure aspect of governance. Learn everything about it. And then lead change on making that thing better.

Your idea has no value. It’s all about execution.

No idea you have today will turn out to be right. It’s not how the world works.

Successful ideas are outcomes, not starting points. It doesn’t matter what your idea is if you’re willing to let your experiences guide you.

Do something. Do it quickly & cheaply. Then see if it works. Get validation. But then have the humility to say some things you did worked better than others. Repeat and expand on what worked. Keep doing that for the rest of your life.


  • If you’re dedicated to fixing democracy, I can’t help you. Your ego is far too big.
  • Narrow your focus. The world is unbelievably complicated. Pick a small problem. It’ll turn out to be a big problem in disguise.
  • Make sure you’re solving a problem real people have, and preferably people with less power than you. Real problems are out there, and you may have to step away from your computer if you want to help fix them.
  • Governance is about power. Power is a social thing, not a technological thing. Websites don’t magically give people power. Your story must be about those your are disempowering too.
  • Find a problem that you can make a unique contribution to, based on your own life experiences. Become an expert in it. Get a real job where you can become that expert. And then lead change on making that thing better.
  • Find the humility to say your idea might be total crap, and demonstrate that you are willing to let your experiences guide you.

None of the ideas here are original, and I am grateful for my friends and colleagues who have kept me centered with respect to issues of power— in particular, Bill Hunt, Laurenellen McCann, and Martha Poon, but also many others.