Bundled Governance
How nonspecific representation drives policy mediocrity
This post on Best Of A Great Lot is a part of a series on the subject of designing a new form of governance. Each piece aims to stand alone, but fits together on the Table of Contents.
In the design of most modern democracies, we pick a single representative to participate in all aspects of policy creation. In many ways it reminds me of a business practice called product bundling, and has a similar set of tradeoffs.
Product bundling
Businesses commonly bundle products together, sometimes because customers want them that way, but often because the business expects it can earn greater profits by selling multiple things at once to a customer who only has to make one decision. Bundles are especially profitable when some of the products have large fixed costs and low delivery costs, so once the product is created, adding it to a bundle incurs little extra cost but earns higher revenue.
The classic example of this is a cable package. Each individual cable channel has a high fixed cost and fairly low marginal delivery cost, so if you can offer the customer an extra dozen channels for an extra 10%, the math works out favorably. Software also follows these rules, so we see so many companies aiming to build platforms, like Google's productivity suite. You can't just buy Gmail for your organization, you have to get their Calendar, Drive, Docs, Sheets, and other such services along with it.
Product bundles offer real value: they reduce the cognitive complexity of having to choose a bunch of different things individually, and for some bundles, allow the business to make them work smoothly together.
But bundling brings with it downsides, too, and sometimes the downsides are sufficient that there's an opportunity for a "great unbundling". Ben Thompson has written a good description of how the television landscape underwent a great unbundling with the creation of streaming services. Josh Breinlinger describes how Craigslist got unbundled with this beautiful graphic:
If you're interested in the dynamics of this, Andreesen Horowitz's blog post on the subject is also worth reading.
At its heart, the tradeoff in bundling is about simplicity - lower cognitive work to interact with something - versus finely tuned control of your choices. Choosing a bundle lets you make a single choice and get many things, whereas choosing individually means more work in the choosing to get the best but more control of what you get.
So it's not surprising that bundling often comes along with a degree of mediocrity, as the whole bundle competes for attention and the individual components can be less than perfect. Cable channels sold as a bundle didn't have to be nearly as good as a streaming service sold separately. Google's office products don't individually have to keep up with the best competing products out there, as most customers buy the whole bundle rather than mixing and matching.
We experience the same tradeoff with our current design of representative government. We don't have to get deep into the policy weeds ourselves, since we only have a few choices, and the winner will be responsible for contributing across all of the possible areas of governmental policy. But our focus on simplicity earns back a dividend of frustration. We're constantly noticing or being told by the news about all of the governance failures that Congress is unable to fix, but then when it comes time to vote for a representative, we don't get to make the best choice for a specific area that is failing, we have to choose among a couple of choices, often choices who don't disagree meaningfully on policy areas that are outside the highest profile areas.
Only some issues are important enough to drive voter behaviors - for many other issues, we are frustrated but unable to gather enough momentum to replace a representative over it.
There are a lot of different aspects of this which combine together to create this situation.
Representatives as policy bundles
With only two (usually) choices for representative and many more areas of policy (let alone specific policies) that the representative might vote on, the math is clear. You're unlikely to fully agree with all of the policy preferences of your top choice, even if you're a partisan conformist at heart. In primaries, you might have a few more choices than that, but rarely more than four serious contenders, so the math remains the same.
Sometimes this might be headline wedge issues like a Democrat who believes in gun rights or a pro-choice Republican, but the mid-tier issues are much more likely candidates for policy disagreements, as they don't have as many activists driving visibility and votes. Even popular policies can be successfully opposed by the majority of representatives if the policy doesn't rise to the level of importance to drive votes.
If you're unlucky enough to be in one of the districts that has only one candidate running or only one serious candidate running, you've got even less choice. If we assume that a margin of victory of 15% is usually because the other candidate never stood a chance, 72% of the House were shoe-ins in 2020.
Writing, not voting, is the real work
Many people focus on the votes of representatives. Did they vote with their party 95% of the time? With the President? How does NARAL or the NRA rate them? I don't want to suggest that this is unimportant, but the focus on it disguises a much more important aspect: the writing of policies. In general, Representatives work on (and have their staff work on) writing policies in only a few areas. Often they're the areas that their committee assignments reflect. Their work writing these policies is often vastly more influential than their voting. The details of policy have a dramatic effect on the outcomes in those areas. Many of these details are built out of a complex intertwining of what ideas the staff, lobbyists, and academics are able to come up with, and what they are able to make politically palatable. That work is rarely as celebrated, but tremendously influential.
Consider the Affordable Care Act. It clocks in around 900 pages, and has spawned thousands of pages of further regulations. Hundreds of choices had to be made about how to frame and write the sections of the law, what aspects of current law it would interact with and how, and how to allow and constrain the Executive branch in its enforcement and follow-on regulation. By comparison, the representatives who voted made a single choice: yes or no. Without their support, the authors of the bill could never have seen their decisions made law, but the larger effect on society is all of the reverberations from the detailed choices the authors made. If we want to improve governmental outcomes, this is where more effort should be spent rather than on getting people to vote the right way.
Or consider Medicare For All. It's a slogan. In 2020 when Medicare For All was a major campaign issue, even the wonkiest of candidates hadn't fully fleshed out what they meant by it, and few of the details even made it into the news. Election campaigns run vastly more on the feelings about candidates than the likelihood of success of their policies.
Who will work on which topics is mostly invisible in elections.
Every representative has to have a campaign opinion about taxes, abortion, healthcare, guns, immigration, but only a few of them will get to do more than vote on those topics.
For the most part, for non-incumbents, we don't have any idea what committee they'll be in, or what areas of policy they'll work on. Only if they have a clear and obvious prior specialty is it obvious where they'll work. Bill Foster, the only physicist in Congress, is in fact on the Research and Technology subcomittee. He's also on the Financial Services committee, a slightly less obvious second choice.
For incumbents, we know what committees they're on, but it's extremely rare for a race to hinge on a representative's prowess writing policy, or what areas they contribute to. Most races are an argument about the state of the country, the headline policy disagreements and votes, wedge issues, and the party itself.
For any particular policy area, the people who care most are often diffuse.
Some topics have enough geographic localization of professional practitioners that you see representatives who are highly responsive to those professionals. Farm states elect farming representatives who work on agricultural issues. AOC is on the Monetary Policy subcommittee of the Financial Services committee because New York is full of bankers.
But most topics don't have this sort of geographic localization. If cybersecurity is the policy area you care about, you're unlikely to live in a place where your representative focuses on it.
Why did I pick cybersecurity? It’s a massive area of governance failure. In 2017, the credit agency Equifax was breached because of common but bad security practices and the records of 150 million Americans were stolen. Despite it being a national scandal at the time, Congress did not manage a meaningful legislative response, and the number and size of breaches has only accelerated. Smaller ones don't even make the news anymore. In 2022, infamous hacker and former security officer at Twitter Peter Zatko testified before Congress that Twitter had terrible security practices. If Twitter can't be serious about security, it seems implausible that many companies succeed at it.
Experts in the industry seem split between frustration that Congress hasn't passed any real legislation and relief.
Most professionals who know about cybersecurity work in one of the bigger tech cities, like San Francisco, New York, Boston, Seattle, Austin, Washington, and a few other places. One might imagine that they would have some representation on the Technology committee of the House, yet in 2022 the only one of those places that has a representative is Silicon Valley.
Is it possible that the representatives on the technology committee happen to come from or have nontraditional means of technological excellence? Given our current legislative reality, it seems a farfetched claim.
Interest groups, lobbyists and corporate influence.
We might not even like it if all of the cybersecurity professionals were in one place, and elected their own representatives to pass professional-approved legislation. I intend to talk a bit more about Public Choice Theory in another post, but a short version of it is obvious in my example of farm representatives above. Farm representatives get agricultural policy passed which is good for farmers, but that doesn't mean that it's good for the rest of us. If you've met as many cybersecurity professionals as I have, you can imagine them colluding to get legislation passed that simply banned email, as it's one of the most persistent threat vectors.
Since mid-tier or below policies like cybersecurity aren't important enough to the electorate to drive their voting behavior, those who care about them attempt to persuade representatives in other ways. It's possible to build a coalition among people who care and then either lobby or apply pressure to get the right policies. For major wedge issues, organizations like NARAL and the NRA compile voting reports on representatives. For more niche issues, work like this usually more resembles thinktanks and lobbying institutions. While they are able to exert influence through a variety of mechanisms, such as campaign donations, offering to write policy position papers and bill language, brokering vote trades between representatives, offering connections and future jobs to staffers, and more, none of these are the intentional design of our system, and when they're made public, they carry at least a whiff of corporate interests and corruption.
Some skillsets are significantly underrepresented in Congress.
This is a little tangential, but still related and often strikes me as interesting given the failures of governance we so regularly see.
Every year it seems another article goes around about how few Congresspeople have a background in STEM. It seems reasonable that if our government is constantly claiming that STEM is the future of our country, maybe Congress should know something about it.
In 2019 Business Insider claimed that Congress had 11 scientists. Perhaps unsurprisingly Business Insider's definition of "scientist" is a little loose.
Computer programmer Jacky Rosen is one of Nevada's two Democratic female senators. The former Congresswoman from the state's 3rd District helped her suburban Las Vegas synagogue install a new solar array. Rosen says that cut the congregation's energy bill by 70%.
Is she a scientist, a computer programmer, or a solar array installer? These aren't actually the same thing, Business Insider. I'll let you be the judge. From Wikipedia:
While working for Summa, she attended Clark County Community College (now the College of Southern Nevada) and received an associate degree in computing and information technology in 1985. She began working for Southwest Gas in 1990 before leaving to open her own consulting business three years later.
But that's only one of the 11. Instead, let's turn back to Wikipedia, because whenever we need a list, Wikipedia is there for us. Of this list, I think it's reasonable to claim either Congress has 1 or 2 professional scientists with doctorates, depending on whether Jerry McNerny's Ph.D. in Mathematics counts. Bill Foster (Ph.D., Physics) is the only uncontestable candidate.
So if you're in Bill Foster's district, you now have two questions you have to ask. Do you think Bill Foster is a good representative, and also do you think Congress is better off having at least one physicist they have to listen to. For the most part, voters don't seem to apply this sort of logic much, but maybe they should. Again, STEM is critical for this country and we have literally one person with a STEM doctorate in Congress.
Even if you're going to count the scientists without Ph.Ds, the number is still very low. The Verge was excited (in 2018) about 7 new scientists out of 435. They stretched the term a bit also, since they include:
A dentist
An industrial engineer who spent a year teaching high school chemistry
A McDonald's franchise owner who once worked as an aerospace engineer.
The Presidency is even more of a bundle.
These days a lot of governing happens in the executive branch, and so it's worth noting that the President is even more of a bundle than Congress. The common phrase in politics is that “personnel is policy”. The President gets to pick the leaders of the major departments of the executive branch, and yet their choices aren't made during the campaign, don't have meaningful input from voters, and while there's lots of rumors of shortlists, don't publicly disclose the reasons or alternatives.
Does it have to be this way?
Despite what it might seem, bundled governance isn't a design constraint. Representative governance is one choice in the goal of self-governance, and having our politicians represent us on all possible issues is another choice on top of that. Other choices are possible.