Bernie Sanders, Diminishing Returns, and the Challenge of American Socialism

Or, In Which I Agree with Ben Studebaker in a Qualified Way

This gets to Ben Studebaker and I significantly agreeing, though we have to meander through a pretty big disagreement first.

Ben puts forward that to get the health care results of Denmark, we just have to scale Denmark’s bureaucracy up. This argument, I will show, is economically naive. But he leaves himself an escape hatch: if we substantially change Sanders’ plan, the many states can probably do what Europe has done. This raises questions about what exactly that would look like, but I am genuinely all ears. This may be the “missing ingredient” in Sanders’ run.

Diminishing Returns

Ben raises the issue of scale this way:

Why would the size of the United States be a meaningful obstacle to the adoption of any social program? I have yet to read a convincing account of why a program that can work for 5 million people cannot work for 50 million or 500 million. The size of the bureaucracy needs to be scaled up, but if we live in a more populous society we have more people available to work in such a bureaucracy.

The answer can be found in any introductory economics textbook. The concept that he forgets is diminishing returns, and there is very little that it does not effect. Now, diminishing returns are often poorly explained in introductory texts and this is because we do not have a foundational model of diminishing returns in the literature. (Wikipedia has a serviceable introduction.) Indeed, most models—from the standard cubic cost equations to the macroeconomic assumptions in the Solow Model—are chosen for mathematical rather than economic reasons.

At any rate, the idea is intuitive enough. As an operation (private or public) grows larger, it must spend more energy and resources managing itself. The difference between providing insurance to 6 million people and to providing insurance to 300 million people is not just 50 fold. As the operation scales, one must hire more managers to coordinate different departments, more specialists to make decisions, and devote more time to lateral coordination.

There are significant advantages to this scaling. Division of labor, specialization, and the resources to make very precise decisions are all major benefits to economies of scale. There is a constant balancing act between accessing the benefits while trying to stave off the inefficiencies. Grow too large, and you pay for it in management costs. Stay too small, and you cannot access the benefits of specialization.

Health Care in the United States

Where health care sits in the USA is not easily discerned. The evidence Ben puts forward is, at face, against his proposition. Medicare is smaller than the major insurance companies, both in scope and number of patients. The expectation has long been that you will have private insurance to chip in for the numerous things Medicare does not cover. This sort of passing the buck is not what I think Sanders has in mind when he proposes socializing health insurance. (I am much more sure it is not what his supporters largely think he means.) Further, Medicare is smaller than the leading firms in the market, who are driving national costs.

It is a bit exuberant to reason from there that Medicare will be past the point of diminishing returns if we extend it up past the levels of those national firms. This evidence is only so much innuendo. My point is much more that Medicare’s effectiveness plausibly lies in its lack of ambition. Diminishing returns offer a good reason to proceed with caution, though not an airtight argument against Sanders’ proposal.

I offer a last piece of innuendo against Ben’s interpretation, but again stress it is little more than that. If economies of scale existed to the magnitude Ben hopes, the market would tend towards monopoly. The largest firm serves, round figures, about one in five people in the United States, or about 13 Denmarks. This is not insignificant, but the has no shortage of competition and is known to have efficiency problems. That gets us to about a fifth of what we need—and diminishing returns implies that the last four-fifths will be even steeper.


A quick word about the Federal exchanges: they are not insurance. They are a marketplace for buying and selling insurance contracts. It is a bit like trying to make an argument about socializing corn farming by talking about supermarket regulation. While changes in how insurance is sold can (and evidently did!) change the insurance market, that is different in virtually every way from having the government run the insurance being sold on those exchanges. It is proof that regulated capitalism can be effective and not an example of any kind of socialism.

Much of the remainder of the evidence in the piece concerns itself with showing that the status quo is stable and is predicated on the idea that there are no diminishing returns. If there are diminishing returns, this too is the argument for the kind of regulated capitalism we have expanded under Obama. It deeply undercuts Sanders’ argument.

States’ Rights

Where Ben and I agree emphatically is that this may well be something the many states should handle. The irony that two liberals would agree on a Federal solution tickles me, but here we are. Europe has arrived at a de facto Federal solution that has worked quite well. I am game for trying it in the States.

But that too undercuts Sanders’ wider project. Pay attention to semantics as I’m about to split a hair: Europe has never tried Democratic Socialism. The many states of Europe certainly have, and it has worked reasonably well. But Federal-level Democratic Socialism has never been attempted, never mind succeeded. (Shots fired: the only European attempt at socialism on the scale Sanders proposes was the USSR.) We have arrived at the case for Sanders to run for Governor of Vermont, preferably with 49 other candidates in other states.

For the sake of space, I must give short shrift to the argument of President Sanders and a favorable Congress creating a mandate for the many states. To deal with it only briefly, to go down that road is to concede that Sanders’ Federal-level approach is flawed. We are then dealing in shades of Federal involvement, which opens up a host of legal, constitutional, and political questions that I do not think get us any closer to dealing with the problems of diminishing returns at the Federal level.

Ben is onto Something

To review, there is a convincing argument for giving diminishing returns a wide berth. In the particular case of health care, the preliminary evidence suggests Medicare’s strength may be in its limited scope, though the issue is far from settled either way. The experience from Europe suggests Sanders’ plan is out of line with the evidence and suggests a state-level plan instead.

And yet, Democratic Socialism survives the diminishing returns critique on the very argument Ben put forward. I throw the ball back in his court in a very different way than I usually do. I am genuinely excited that we may be on the cusp of agreeing on a sort of Federal socialism. Long experience tells me that we may not agree—but we will see.

The obvious question that presents itself is this: What does a Sanders revolution look like if we accept diminishing returns effect some of his proposals? (Also: what are Ben’s thoughts on diminishing returns?)

I have some thoughts, but I’d genuinely like to hear from Ben. He has changed my mind already to think the states are much more capable of socialism than I thought prior to reading his piece, and I would like to see where he traces the thoughts that changed my mind.


Reality Just Called: Bernie Sanders Did not Win the Democratic Debate

I am seeing a flood of Facebook statuses that cover the usual range of reactions to something happening involving Bernie Sanders. Why is “the media” acting like Clinton won? (She did) Why are polls not being released? (It takes about this long.) There must be some connection between these media companies and Clinton. (It is complicated; yes, but several polling firms we’re waiting on are independent of the Clintons.)

One of the first “flash polls“, polls that are taken quickly over a single day to give news outlets a snapshot to frame the debate, is out. It found that Clinton won. And it was not close. Of the 760 people they asked, 60% said they thought Clinton won the debate. Chaffee and Webb were deemed to be the losers by a long margin.

Fans of Sanders need not despair here. Both Clinton and Sanders were viewed more favorably after the debate. Perhaps more importantly, Sanders is behind Clinton just outside of the margin of error. Clinton maintains a lead, but it is hardly safe. This cuts every which way: Sanders is narrowing the margin, Clinton has not been overcome, some Sanders supporters evidently think Clinton bested him in spite of their personal views.

This poll differs from most proceeding polls in that it takes Biden out of the picture. Sanders seems to have snapped up more of Biden’s support than Clinton. I find this exasperating, but it is easily explained. The Vice President is as much a political insider as anyone, but he has always managed to project an image of being an outsider. The differences in a Clinton and Biden presidency would be in competence, not policy goals. Lest I leave any doubt here, Clinton is the more competent candidate by leaps and bounds.

The usual caveats apply: this is a single poll taken on a single day. You can only extrapolate so much and it will be a few days before a clearer picture emerges. It is possible the first poll out the gate is an outlier.

But guys. Sanders did not apparently win the Democratic Debate. He performed well and energized his base. But most registered democrats felt Clinton won. Not Sanders.

A Pack of Memes

The internet continues to have memes which are wrong and in need of correction. I continue to be happy to offer my services!



I do not think the author of this gets how religion works. It is totally, completely within reason to say that if something is an affront to God, it should not happen. It is a short skip to making it illegal. Plenty of religions take this line, and in fact, it is pretty typical for this happen when the majority of stakeholders in a government follow the same religion. Even in the relatively secular laws of Western Europe and the United States, Christian theology and Christian faith animate the legislative process. This is fairly normal; if you elect a bunch of people with deeply held convictions, they will legislate those convictions.

I’m not against saying that we should consider this meme a civil ideal. But if people do not understand what is here, they do not understand how the law works. That is completely different. Religious law may, for example, forbid you to watch Doctor Who on Sundays; secular law requires that we not ban the airing of Doctor Who. (I look at Indiana, which bans the sale of alcohol on Sundays. I glare disapprovingly and move on.) But people may have a religious conviction and duty to make society in their God’s image, and they understand that theology perfectly well.

Ten Commandments of Logic

Ten Commandments of Logic


Some fallacies are formal fallacies; they are shaped wrong. You cannot say, for example, that A therefore B and C therefore D implies A therefore D. This is the four term fallacy, and it occurs most often when B and C are subtly different or homophones. This kind of logic will never prove the conclusion. (The conclusion may, of course, be true through different logic.)

These are all informal fallacies. That means that they are content that is suspect. They are not wrong at face, however, and may represent imperfect heuristics for dealing with incomplete information or be appropriate in certain circumstance.

  1. Ad Hominem: One of the stronger informal fallacies, ad hominem can only be useful when the person’s character is material. In most arguments, bringing up a person’s supposed faults is a kind of derailing or failure to engage with the argument. But it can be useful in setting burdens; if someone is known to write falsely about a topic, it is worth starting there to direct scrutiny. But Shakespeare has this one in the end—“But soft! Does the Devil speak true?”
  2. Straw Man Fallacy: The straw man fallacy is one of the informal “fallacies” that is not really a fallacy per se. As long as the argument is done properly, it should be nominally correct. If something is a straw man, however, it fails to address some critical part of the original argument. It simply fails to clash with what came before it, even if it is true on its own terms. The straw man can be a useful rhetorical device—but proceed with caution. It should take the form, “This argument is similar to [weaker argument] which is incorrect, but it is different. Still, those differences do not negate the essential problem.” Thoroughly explained, it is arguably not a true straw man. Regardless, applying this commandment dogmatically eliminates a valuable rhetorical tact.
  3. Hasty Generalization: A full detailing of this is called a “statistics textbook”, but the essence is easy enough. Small depends very much on the topic at hand and some a priori assumptions we make about the sample. The bottom line is that the strongest forms of logic make no generalizations; induction is considered inferior to deduction.
  4. Begging the Question: Begging the question is also not a fallacy in a strict sense. It is just a case where you put down a tautology, something that is true by construction. This can be very powerful as a starting point or for correcting an analysis that abuses technical terms. It is also never wrong, it just does not get beyond the level of definition.
  5. Post Hoc Ergo Propter Hoc: Things are complicated here. This is a strong informal fallacy in formal logic, but a weaker one in empirical science. In formal logic, we should be hostile to cause claims from the order. But when paired with a good reason, it can be moderately useful for suggesting an empirical model. Correlation may not imply causation, but we should probably look closer if we find correlation.
  6. False Dichotomy: The false does a lot of work in this one. Dichotomization is a powerful rhetorical tool, and can often illuminate a complicated problem if you can find two poles. But if the poles are bad representations or exclude either the middle or another choice, then it is less useful. The burden is on the person pointing out the dichotomy to point not at the existence of two poles, but rather find the unrepresented part and show it matters.
  7. Ad Ignorantum: Also a fairly strong informal fallacy. It is very rarely correct to point to a lack of information as anything but proof of lack of information. In fact, this one usually gets misused when we have a strong set of negative findings. Having not found, despite looking into a number of miracles, artifacts, and other reports evidence of a God or Gods is evidence of negative findings, not lack of information. (Caveat: there is a fairly clever way around this by pointing out that God might not be beholding to empirical science. Does not change the fact that we do not have evidence.)
  8. Burden of Proof: I liked this one enough to name my blog after it, but tread lightly. This applies to the question, “What proof do you have?” It is wrong to answer “What proof do you have?” (Do notice, this is largely a rehashing of ad ignorantum.) But it is completely, totally within reason to ask them to address your burden. “Where have I failed?” The burden of proof of a person raising a counter claim, “No, something else is correct,” is equal to that of the original claim. There is no benefit to being the second person to make a claim. This is probably the only one on here that might be a commandment, and it still gets abused.
  9. Non Sequitur: Another instance where you should tread lightly. The non sequitur is always wrong if judged a non sequitur, but it is always subjective as to whether or not the parts flow from each other. Sometimes it will be obvious. Sometimes it will be open to debate.
  10. Bandwagon Fallacy: This one is another one that is strictly true, but can be useful nonetheless. Stories of brave people standing up against a popular position that was later discarded are dime a dozen. But consensus is also a powerful tool, especially if the people who make up the pool of people coming to a consensus have something else go for them. It is technically an informal fallacy to point out that the overwhelming majority of climate scientists believe in global warming. It is also valuable information—people who consider the problem in a competitive environment overwhelmingly agree. With no other evidence, it is a poor argument, but it is raised in the context of other arguments, it is persuasive.

The essence here is that informal fallacies are complicated and subjective. As far as a Ten Commandments of Logic go, this is a poor offering that misses what makes logic tick.



With our new understanding of informal fallacies, I now give you e being a bit more of a hardliner about them. This is a useful thing to keep in mind—and a Marcus Aurelius quote for the win! But this actually the opposite of the false dichotomy. It is an informal fallacy to suggest that just because there are two perspectives, the truth must lie between them. This is a very, very useful heuristic, but some things really are polar. Some arguments really are wrong. I 100% endorse checking for the excluded middle if you hear two arguments, but I’m not guaranteeing you’ll find it. At all, in any way.

Social Security

Social Security

This last meme is one of those things that is true but a bit misleading. Social Security is in fact accounted for as a trust. Legally speaking, workers who put money in are owed money back.

But saying that social security is funded by American workers misses elements of the whole game. Who pays income tax? Where, pray tell, does most of the money that goes to sales tax start from? What the Federal Government did was borrow money against their trust books. The trust still has plenty of funds, but it is in the form of illiquid bonds. The government could at any moment make social security liquid again by buying back its own bonds from the trust. In fact, it will have to at some point.

That is important, but a lot gets made out of the fact that social security is solvent because we could make other parts of the government insolvent. It is a shell game on every side. Sure, the GOP is being duplicitous (perhaps unknowingly) by suggesting social security is going to run out of money. But the point that there is not quite enough taxes to go around remains when liberals suggest paying back social security.

You’re welcome internet, you can now be a little more correct!

A Note on Angus Deaton’s Nobel Prize

I am finally settled into my new job in a place that is very, very far from where I grew up, and ready to pick up blogging again. And what better time to return than with the Nobel Prize in Economics?

The thing to understand about the prize is that it is as a rule given to work that is reasonably well-known but is deemed to be deserving of more attention. I had read Deaton’s popular work, The Great Escape and cited it in my senior thesis in economics, but I had to check my works cited to be sure of that. This is the committee’s way of saying you should have read it too, and that people in econ should pay more attention to his technical papers. (Guess what I’ll be reading in the weeks to come, because I’m a ginormous nerd!)

I suspect that Deaton was selected in order to combat the focus on macroeconomics that has followed The Great Recession. Deaton is very critical of aggregation on the grounds that it tends to get rid of details that can tell you why individual circumstances vary. The rush to look at GDP and market trends following the downturn has been less careful about this. Even Picketty’s Capital in the 21st Century accepted aggregation as a premise. The committee is reminding us to look at the details of circumstance driving poverty, inequality, and wealth.

A cursory look at Deaton’s work makes me think he is a bit zealous—aggregation can be useful in its simplicity. He likely recognizes this in a wider sense. There has been a tendency to exaggerate the claims of Nobel winners. Last years winners proved that market bubbles do not exist in a certain, well-defined sense and yet their findings were covered as if they argued that bubbles did not exist at all. (This was compounded by sloppy interviews by the winners, but economists do not get Ph.D.s for their public speaking.) Deaton’s skepticism of aggregation and the policies it recommends—aid, stimulus, and welfare—deserve scrutiny. But like every winner, this will include deciding against some of his ideas.

What we have is a good example of a pretty standard Nobel win: an economist doing solid work with a worthwhile if not airtight theme that deserves to be taken more seriously.