This is a cross-post from the blog of the Centre for the Study of Governance & Society at King’s College London.
Over the last two decades online services have transformed from a product of a multitude of enterprises to being dominated by a handful of corporate-owned platforms such as Apple, Microsoft, Facebook, Google and Amazon. They specialize in connecting media producers to users. These are often mutual interactions with users both producing and consuming content. These platforms play an increasing role governing commercial exchange, as well as civil discussion, with plausibly pernicious implications for liberal democracy. As I propose in a recent paper ‘Markets for Rules’, blockchains offer a promising solution to this danger by helping to displace corporate ownership in favor of common platforms sustained by users themselves.
Corporate concentration has produced enormous efficiencies and innovations, improving user experiences and boosting investment in hardware and infrastructure. But it has also had several bad consequences. These enterprises face extremely low marginal costs and network effects whereby additional users add value to an existing user-base. Some of these effects are explained by these platforms’ business models of collecting personal data to target advertising more effectively at customers. The more interactions on a single platform users have with each other, the more useful the data for advertisers. The result is overwhelming returns to scale and a winner-takes-all competition for profits.
This has troubling implications for economic inequality, especially if we end up with a handful of corporations taking a bite out of every conceivable transaction. Of greater concern is the way owners exert control over who can join and what people are allowed to do on their platforms. Content producers can be demonetized or banned, effectively denying them access to a user-base or revenue. Online sellers can find themselves frozen out of a platform payment system without legal remedies. Controversial or unpopular producers survive at the whim of executives or, at best, a patchily enforced official policy.
This reliance on private governance is a problem for consumers, producers and ultimately citizens. But it is also a challenge for executives who find themselves mediating acrimonious personal disputes and political debate. With all the data in the world, they struggle to judge consistently what belongs on their platforms. The fact that these corporations have ended up functioning as unofficial censors and wielders of sanctions has led some commentators to propose regulating these platforms as public utilities or, more radically, nationalizing them so that access to them is decided democratically. These solutions have their own perils because any centralized system of monopoly control, whatever the underlying democratic credentials, can produce authoritarian outcomes. Liberal democracies up until now have been sustained by an independent civil society constituted by overlapping and competing spheres of governance, not the monopoly of either democratic or corporate government.
The prosecution of the CEO and founders of Backpage, who failed to exclude sex workers from their platform, illustrates the reliance of these private enterprises on government support on controversial policy issues even in relatively free societies. The combination of privately-developed data-collecting networks with over-arching state control is arguably reaching a nadir in China which is rolling out an unaccountable surveillance system of ‘social credit’ that can identify political dissidents and automatically exclude them from significant spheres of civil society.
Is there a way that blockchains can help navigate around the centralising and authoritarian impetus of technology-facilitated governance? Blockchains emerged from two pre-existing technologies – public ledgers and asymmetric cryptography – to produce a way of sharing data across a network that is resistant to manipulation by unauthorized actors. Initially conceived as offering alternatives to state-backed currencies, blockchains are now used to build decentralized autonomous organizations (DAOs) and dapps (decentralized apps). They can supply similar functions as corporate platforms but without an overall owner.
These systems are sustained by rewarding network participants with tokens (through completing intensive computing processes called mining). Tokens are convertible into ordinary currency, albeit currently at volatile rates. The entrepreneurs that build these platforms typically reward themselves and investors a large stake in those tokens but once the network is launched, they do not have control over how it is utilized. The rules of each network are self-enforcing. These rules can be changed, either through the original (or new) developers launching a rule-set that others may choose to switch over to (a fork). Alternatively, the rule-sets might contain provision for amendment. Such amendment schemes are, of course, open to manipulation as is the case for all political processes. Nevertheless, what these schemes offer is a way of interacting and exchanging at large distances without an overarching ruler. Instead, conduct is permitted on the basis of fixed rules enforced mechanically by people’s decisions to participate in the system. One way of looking at these schemes is that they have decentralized properties of communal norms, combined with the possibility of more deliberate design and experimentation of more formal rules and institutions. I call this common government.
The implications of this new technology and kind of governance might turn out to be very far-reaching, approaching that of the development of the Internet itself or even the printing press. But what could it mean for familiar Internet platforms in the medium-term? First, participating in mutual platforms might better align the incentives of users and platform designers. Right now, platform owners rely on squeezing as much data out of users as possible in order to sell it on to advertisers and to sell additional services. Mutual platforms, without responsibilities to shareholders, can experiment with different funding models. Individual users might elect to sell access to their profile to advertisers but the data itself can be made more secure as it will be a property of an encrypted network rather than a profile stored in a central private database. Privacy can be better assured than private management with public regulation.
Second, the networks can be more robust both to natural and political perturbations. Under decentralized protocols, ordinary users help store and serve content to each other. With the addition of blockchains, these users can be compensated for making their idle computer resources available for network use. This means that data doesn’t have to travel so far as is currently the case from host to user and the network as a whole can better cope with outages from particular nodes without data loss. Without a central controller, there is no particular agent that a government can coerce or punish for allowing specific interactions over a platform. Governments would then face the more difficult choice of permitting or prohibiting Internet communications altogether. It is thus more robust against arbitrary government censorship and manipulation of trade.
The relationship between users on a platform is mutual. The relationship between users and platform owners, however, is presently hierarchical – a private dynamic that government agencies can exploit. What blockchains may eventually permit is the provision of relatively efficient networks reliant neither on a single public agency nor private owner.
Learn more about Nick’s work here.
I recently had a paper rejected in Political Analysis. I fully expected a rejection given the journal’s high ranking and had submitted it for the sake of feedback. As I’m sure academic readers know, getting someone to read our papers can be hard. Unsurprisingly I got a rejection notice earlier today.
Surprisingly, all reviews were actually constructive feedback and, while critical, kind. To add strangeness to the whole ordeal, the process took less than a month from submission.
Since I have no way to contact the anonymous reviewers, I post my thanks here. Thank you anonymous reviewers. I wish reviewer #2, and here too I’m sure academics know the type I am referring to, should be more like you.
Brexit isn’t the only ridiculous thing happening in the United Kingdom. In April, the British government is rolling out statutory adult verification for pornography websites and content platforms. This requires all adult content providers to have proof of age or identity for all their users, whether a passport or a credit card (or more ludicrously a ‘porn pass’ that Brits wishing to browse anonymously will have to buy from local newsagents). The government plans to require internet service providers to block pornography websites that are not in compliance with adult verification once the system is in place. For those with university institutional access, Pandora Blake has written a timely explanation and critique published in Porn Studies: ‘Age verification for online porn: more harm than good?’.
Technical challenges with rolling out the system have led the dominant pornography search platform owner, MindGeek, to develop proprietary solution, AgeID, in cooperation with regulators. This cooperation between the dominant commercial pornography platform supplier and a Conservative government publicly intent on restricting access to pornography might appear surprising. However, it can be explained by a particular pattern of regulatory capture identified in public choice theory as a Bootlegger and Baptist coalition. Bruce Yandle observed that throughout the 20th century, evangelical Christians in the United States agitated for local restrictions on the sale of alcohol with the avowed aim of reducing consumption but with the secondary effect of increasing demand for alcohol for illegal bootleggers. Hence both interest groups, apparently opposed in moral principle came to benefit in practice. We now have a classic British case study. In this case, MindGeek is not acting as a literal bootlegger. It intends to be fully legally compliant with the filtering regime. However, the law will block all non-compliant competitors without a comparable verification system. They can gain a competitive advantage with a proprietary technical solution to the barrier introduced by the government.
Introducing identity verification systems has high fixed costs and low marginal costs. It is costly to develop or implement but easy to scale once integrated. The larger the pornography enterprise, the more easily these costs can be absorbed without the risk that it will not be worthwhile to serve the British market. For many smaller international pornography websites, without in-house legal advice or technical expertise, it might prove uneconomical to serve British users directly. So MindGeek’s platforms could become the least-cost legal gatekeeper between small enterprises producing pornographic content and the British public. The government is raising transaction costs to accessing pornography in a way that impacts larger and smaller platforms asymmetrically and favors one dominant platform in particular.
Both the premise of this policy and its likely impact on the market for pornography is unpromising. At its most benign, this could be a characterized as a ‘nudge’ against the consumption of pornography and reducing access of inappropriate content to minors. But these limited benefits have costs for both producers and consumers. On the consumption side, it increases risks to data security and privacy because it will plausibly tie records of pornographic access to verified identities, with a clear likelihood of being to infer an individual’s sexuality from private browsing. This could represent a particular vulnerability for LGBTQ identifying individuals who live in communities where there is still stigma attached to minority sexual orientations.
On the supplier side, it takes what already appears to be a market with strong tendencies towards a winner-takes-all model, and then augments it so that a dominant platform has a legally enforceable competitive advantage over potential rivals in the market. Ultimately, it threatens to further strengthen the bargaining position of a single corporate pornography platform against the sex workers who supply their content.
This post is in response to feedback from my previous post on this topic.
There are no panaceas.
But as abstract ideas go, pollution taxes are pretty appealing. Holding constant lots of things that we can’t really hold constant, it means replacing the inefficiency resulting from poorly defined/enforced property rights with a world where prices more accurately reflect the costs of one’s decisions.
Let me come back to the things we’re “holding” constant in a bit. Why do I want to throw my weight behind shifting public perceptions in favor of pollution taxes?
I think they’re underrated by the median voter. Climate change is just a subsidy paid in the form of worse conditions. But most people (including people who should know better) don’t have a good understanding of the problems caused by subsidies.
Which is not to say a carbon tax isn’t overrated by the median policy wonk. There are a ton of important caveats, but on balance, as a policy for use in the next 50 years, I think they’re a useful tool to enhance efficiency or replace worse tools.
Again, there are no panaceas. I’m also not a huge fan of the “Economists’ Statement on Carbon Dividends” as written (for reasons I’ve hopefully mostly addressed). I suspect the best case scenario for my preferred carbon tax policy would be a modest improvement. I think the bulk of the gain would be a cultural shift away from “let’s regulate our problems!” to “let’s leverage incentives to address our problems!” Not Earth shattering, but a step in the right direction.
So let me state my position, then we can dig into criticisms and caveats.
Let’s make marginal shifts away from taxing investment and towards taxing negative externalities. As we go, let’s spend a lot of effort trying to study the impacts and adjust accordingly. Let’s heavily agument that with abatement policies rather than trying to return to some pre-industrial climate target.
Okay, let’s dig into criticisms and caveats.
- Public choice considerations
- Geoengineering and other alternatives
1-Public choice considerations
A Green New Deal will be a rent-seeking bonanza. Pollution taxes will face the same sorts of problems that plague the tax code in general. There will be intentional loop-holes and accidental screw ups.
We have to continue to push for reducing the complexity of tax codes in general. But I can’t deny that a carbon tax would be a step back on this margin.
Minus a hundred points for my position.
2-What about geoengineering?
Geoengineering sounds like a possible panacea. Maybe it is. But I’m not willing to flip a switch and find out the hard way all at once.
First off, geoengineering is scary. The climate is a complex system and complex systems are difficult-impossible to manage well. And that’s especially concerning if it means that anyone with a few million bucks can try to fiddle with Earth’s thermostat.
But it seems like a plausible tool that might be used to address climate change. Similar to my take on a carbon tax, I think the way to go is baby-steps plus research.
What about subsidizing “green _____”
Personally, I’m skeptical. Solar sounds appealing, and I (personally) think windmills are beautiful. But I don’t think the government will do a good job of picking winners and losers. Pollution taxes are appealing to me because they don’t require bureaucrats to choose. Again, I think the way to go is to use pollution taxes to offset other taxes–while continuing to advocate for reduced size/scope of government and a return to federalism.
Plus five points for my position.
We should also remember that GDP is an imperfect measure of well being. The current figures aren’t directly comparable to the figures we’d get in a post-carbon-tax world. A one-time fall in GDP doesn’t (necessarily) mean we’ve screwed things up.
A tax big enough to halt climate change would be incredibly costly. Too big a tax yields a negative net benefit.
Still, it’s worth remembering that a) we can go too far with a carbon tax, and b) we don’t have access to a silver-bullet solution. So let’s start small and gradually increase carbon taxes till we get close to (our best estimate of) the optimal level.
Plus epsilon points for my position.
The basic idea of a carbon tax is that we’re dealing with a global-scale externality problem. But small scale taxes are unlikely to do much beyond shifting where pollution happens. A fully effective tax would require multi-lateral coordination. And, as a country, we aren’t very good at that.
Trying to create a tax on imported carbon-intensive goods that didn’t face a tax at home seems a) sensible at first blush, and b) a massive opportunity for public choice problems.
On the other hand, we could justify a tax commensurate with the local impacts (something like 10% of the global impact). This fits nicely with my idea of starting small and adjusting at the margin.
But even within the U.S. there are coordination issues. Long Island will likely face net costs from climate change, but other areas will benefit from a longer growing season.
Plus 10 points for my position, but also minus 10 points.
Uncertainty cuts both ways: we’re currently accidentally manipulating the climate and that could turn out to be catastrophic. Trying to intentionally manipulate it in the other direction is also dangerous. Again, the appropriate focus is on marginal tinkering [much as it clashes with my non-interventionist priors] rather than ambitious global engineering [which grabs my priors by the lapels and knees them in the groin].
When I teach externalities, I draw a graph like this:
In this market, we end up with an equilibrium quantity defined by the point where Marginal Private Cost equals Marginal Social Benefit (MPC = MSB). But the Marginal Social Cost (MSC) is greater, so we get a deadweight loss equal to the triangle I’ve shaded in red and purple.
It’s important to note: we don’t actually know where the MSC curve is. It’s somewhere above MPC, but we’re basically in the position of trying to eliminate a subsidy we don’t know the size of.
The relevant models–climate models and economic models–are filled with uncertainty that we simply cannot resolve without real life experience.
What does the economic way of thinking tell us? Act on the margin. Setting a tax that pushes supply (MPC) up to the green line doesn’t fully address the problem (as I’ve assumed it to be in this graph), but it’s an improvement.
Even better, it’s an improvement where the biggest returns are experienced up front. This modest tax fails to get rid of the red deadweight loss (DWL) area, but it eliminated 3/4 of the total DWL.
Plus X points for my position where X is a random variable with an unknown distribution, positive first derivative, and negative second derivative.
At my friend’s behest I’ve been looking at Bob Murphy’s critique of carbon taxes. I find it’s shifted the magnitude of my prior opinion, but not the direction. I still think carbon/pollution taxes are a good idea, but I no longer think they’re a great idea. My take away from Murphy’s work is that the optimal carbon tax is fairly modest. My response is to advocate for getting a very modest carbon tax on the books, then gradually shift tax policy in that direction.
For climate change (and any other problem) we ought to be pluralists. A mix of approaches is ideal. Part of the appeal of Pigouvian taxes is that they allow and encourage a wide range of responses. The best pollution abatement scheme isn’t something we can look up in a binder. We have to discover it, and crowdsourcing is the appropriate way to do that.
But carbon taxes are only one part. We should also advocate for changes that will ameliorate harm. I am more bullish on these policies than I am on a carbon tax:
- Make it easier for the world’s poorest people to move to rich countries that will be better able to cope with climate change.
- Quit subsidizing flood insurance.
- Quit subsidizing polluting industries (and other industries).
Even though geoengineering scares me, we should try to learn more. Ditto for any other possible tools that come along.
Neomi Rao, nominee to the D.C. Circuit Court of Appeals, has fallen victim to an old fissure–conservative and libertarian disagreement over unenumerated rights. At the hint that she believes that rights exist outside the express text of the Constitution, conservatives have pounced, concerned that Rao will betray conservatives on abortion issues. On both political and legal dimensions, this concern is silly.
As a general matter, it’s beyond clear that Rao is qualified. Conservatives shouldn’t cripple a great candidate over a minor issue. The Supreme Court has many times confirmed that unenumerated rights exist. Rao is bound by those precedents, including Roe, and whether she agrees with those precedents or not is immaterial to the job she’s nominated to do. The chance that Rao will even have a chance to expand the existing list of recognized unenumerated rights is exceedingly low. It just doesn’t come up that often, and the courts already have tests for assessing whether a right should be recognized.
But perhaps more importantly, non-enumerated rights don’t lean toward one side of the ideological spectrum or the other. The Supreme Court of the early twentieth century recognized, for instance, liberty of contract as a constitutionally protected right–though unenumerated. Progressive jurists bent on defending Roosevelt’s New Deal did so in part by opposing the enforcement of unenumerated rights. Unenumerated rights also include other “conservative” causes such as the right to earn a living. And, of course, unenumerated rights have also favored “liberal” values such as a broad right to privacy. In short, it is unclear why Rao’s alleged support for the enforcement of unenumerated rights should enter into the partisan calculus.
And then there’s the simple fact that unenumerated rights do in fact exist and deserve constitutional protection. The history and passage of the Ninth Amendment, which says enumeration of rights shouldn’t detract from those retained by the people, make this clear. The founding generation didn’t see rights as reserved to a fixed set of especially important activities. In fact, many feared that the enumeration of certain rights would imply that the unenumerated ones shouldn’t be recognized. That fear has turned to be prescient, despite the inclusion of the Ninth Amendment, which was written to make clear that the enumerated rights shouldn’t be seen as implying that unenumerated rights should go unprotected.
In short, not only would Rao be right to recognize such rights–though the scope of those rights is always a matter of intense debate–she has to recognize them under binding law, and her thoughts on the matter are unlikely to make much difference to her job. If she gets nominated to the Supreme Court later down the road, the concern may have more relevance. For now, just get her confirmed.
Last election, Jillian Ostrewich drove to a polling place at a Houston rec center, expecting to vote. But she made the mistake of wearing a “Houston Fire Fighters” t-shirt. An election worker confronted Jillian, insisting that she couldn’t vote unless she turned the offending shirt inside out. The rationale: a measure on the ballot was related to firefighter pay, and the shirt was related to firefighters.
Tony Ortiz suffered a similar fate in Dallas. Tony’s crime was wearing a MAGA hat while he stood in the voting line outside his local library. An election worker said he couldn’t wear the hat to the polls because the MAGA slogan constituted “electioneering.” Tony responded that the hat had nothing to do with any issue or candidate on the ballot. The election worker threatened to call the police.
These election workers were relying on a Texas law that forbids electioneering or wearing “a badge, insignia, emblem, or other similar communicative device relating to a candidate, measure, or political party appearing on the ballot” in or near a polling place. The law even prohibits someone from wearing an ID if the name is the same as someone appearing on the ballot. A violation is a criminal misdemeanor. Today, my colleagues and I at Pacific Legal Foundation filed a lawsuit on behalf of Jillian and Tony challenging the Texas ban.
As it happens, the Supreme Court issued a decision just last year striking down a similar law in another case brought by Pacific Legal Foundation called Minnesota Voters Alliance v. Mansky. The Minnesota law in MVA prohibited wearing a “political badge, political button, or other political insignia” in the polling place. The Court said states can limit some electioneering at the polling place to prevent voter intimidation and excessive disruption. But the government has to have some clear and logical basis for sifting “what may come in from what must stay out.”
The word “political” was not a clear boundary. As the Court noted, almost anything could be considered political, and the fuzzy language offered too much wiggle room for abuse and discrimination. The attorney arguing for Minnesota drove this point home during an onslaught of blistering questions in oral argument: how about a rainbow flag shirt? Permitted, maybe. How about a “Parkland Strong” shirt? Permitted, probably. How about an NRA shirt? Definite no. How about a shirt with the text of the Second Amendment? Definite no. At about that point, most everyone knew the outcome of the case–the attorney had just proven the inherently arbitrary standard he was trying to defend.
The Texas case filed today is a bit different. For one, the ban extends to 100 feet outside the polling place, which clearly encompasses sidewalks and other public areas where speech rights enjoy their widest berth. But the ban is not quite as amorphous as the word “political.” Instead, the ban extends to messages that relate to a candidate, measure, or party on the ballot.
This slightly narrower language probably doesn’t save the Texas law. After all, almost anything can be considered related to a candidate if the candidate has taken any kind of position on it. If a candidate has criticized Trump (have any not?), then is Tony’s MAGA hat “related” to the candidate? If Ben & Jerry’s takes a position on a ballot measure, does a Ben & Jerry’s shirt relate to that ballot measure? Maybe, but it depends on the election worker–and therein lies the problem. As the Supreme Court said in MVA, “A rule whose fair enforcement requires an election judge to maintain a mental index of the platforms and positions over every candidate and party on the ballot is not reasonable.” Not only would such an index be impossible, but the contents and application of that index would be dictated by an election worker’s own biases and background.
The integrity of the electoral process is vital. But surely Jillian’s firefighter shirt or even the much-reviled MAGA emblem do not imperil the right to vote. Wed don’t shed the right to express ourselves in peaceable ways when we step into a polling line.