Augustus – G.O.A.T?

2000 years ago today, Caesar Augustus died.  That’s right, in the month named after him.  More or less how you know you’ve arrived.  Since Augustus both 1) won a brutal civil war and 2) behaved fairly well afterwards, he is remembered as one of the greatest leaders of all time rather than one of Rome’s many warlords, pretenders, and brutal tyrants.

His status of Greatest Of All Time (G.O.A.T.) is hardly unquestionable, however.  He is rightly recognized for ending the civil wars that had wracked the Roman Republic and putting the Empire on a much firmer foundation, which lasted peacefully for two hundred years, more erratically for two hundred more, and as the Byzantine Empire for a thousand years after that.  However, the institutions Augustus left behind were not unquestionably better.

The institutions of the Classical World were hardly perfect, but were better than the Medieval ones that displaced them.  Classical institutions were generally legalistic and governed by impersonal rule – e.g., the Athenian assembly and the Roman Senate.  The Roman civil code was a masterpiece that is the ultimate root for Western legal systems, including accountability for political leaders and protection of private property.  It also included slavery, but hey, no one’s perfect.  This legalistic and impersonal rule is generally taken to be a prime prerequisite for stable government and economic growth in the modern world.  Augustus’s takeover supplanted impersonal legalistic rule with personal autocratic rule, which incentivizes instability by making it much more rewarding to get to the top of the heap.  Think the Game of Thrones – “you win or you die”.  Personal leadership and its attendant institutional instability dogged the Western World from Augustus until the Early Modern Era (16th/17th century) and much longer in some places.

Might Augustus be responsible for sending the West on a 1500-year-long journey down a developmental dead end?  That would probably qualify him as the W.O.A.T.

Why Rent Control Will Never Die

I have relatively little to add to a brilliant post by Pedestrian Observations (a fantastic urbanism blog) about the economically and politically toxic nature of rent-control regimes, and the difficulties in transitioning away from them.  It’s worth excerpting two big chunks.  The first is a good explanation for the path dependency rent regulation creates:

As the gap between the regulated and market rent grows, landlords have a greater incentive to harass regulated tenants into leaving. This is routine in New York and San Francisco. Community groups respond by attacking such harassment individually, which amounts to supporting additional tenant protections. In California, this is the debate over the Ellis Act. The present housing shortages are such that supporting measures that would lower the market rent has no visible short-term benefits, and may even backfire, if a small rent-controlled building is replaced by a large unregulated building.

 Here is what is in my mind the key section for understanding the deeper reason why rent regulation battles rip cities apart:

Instead [of market pricing], cities give preference to people who have lived in them for the longest time. Rent control, which limits the increase in annual rent, is one way to do this. City-states, i.e. Singapore and Monaco, have citizenship preference for public housing to keep rents downfor their citizens. Other cities use regulations, including rent control but also assorted protections for tenants from eviction, to establish this preference. Instead of market pricing allocation, there is allocation based on a social hierarchy, depending on political connections and how long one has lived in the city. People who moved to San Francisco eight years ago, at age 23, organize to make it harder for other people to move to the city at this age today.

I’d simply add that this is a classic situation of Polyanian fear of the market.  Polanyi’s great insight was that the “free market” was a utopian dream – when left to their own devices, people rarely self-organize into market provision for crucial goods.  In fact, when the state intervenes to attempt to bring a market where traditionally goods where provided socially, it generally provokes a violent backlash.  Rent control is a great example of this – it creates two parallel systems, one a free-market system and the other where apartments are provided via a social hierarchy primarily based on who was there first.   Ending rent control not only can raise prices for the current insiders, but much more threateningly erases a market of their social status.  I hadn’t quite ever thought of rent control as “provision via social hierarchy”, but in that light it makes it much more clear the personal rage that the subject inspires.

Anyway, read the whole thing.

The GOP’s Uber Gambit: A Theoretical Interpretation

Like Yglesias, I noticed that Republicans have been trying to politicize Uber.  I thought it was pretty neat, in large part because this is one of the cases where political science gives us pretty clear predictions.  This effort won’t work, but even if it did it would likely be bad for both the Republican Party and Uber.  First, let me explain why this is unlikely to go anywhere:

In order for an issue to become a partisan issue, it must be both salient and divisive.  Salience just means whether an issue is important enough to be near the top-of-mind for many Americans.  The economy, welfare, crime, (maybe) foreign policy; all of these are persistent salient issues for Americans.  Others become more salient at specific times – the environment, civil rights, gay rights, etc.  The mechanisms of how issues become salient is complicated, but let’s ignore that for the moment.  Uber just isn’t important enough to ever become a salient issue for most Americans – especially since it only operates in large urban areas where most Americans don’t live.  Ah, you might retort – but that’s where young voters the GOP is trying to win live!

Sure, but it fails the test of divisiveness.  Most Americans don’t see any reason why Uber should be illegal*.  In fact, outside of taxi lobbyists and trade associations, I suspect you will have trouble finding any meaningful number of Americans opposed to allowing Uber.  If the Republicans succeed in raising the salience of Uber to any degree, the Democrats will side with Uber as well.  The current state of Democrats passing anti-Uber laws is an oddity simply because legislators protect incumbents when allowed to operate without public scrutiny (i.e., low salience) and because legislators in large cities are virtually all Democrats.  If the issue became widely salient, the Democrats would not obligingly line up behind a massively unpopular policy stance.

The implicit theory here, by the way, is that Uber could drive a realignment wherein the GOP captures younger voters.  The idea here is that the public has a wide range of opinions along multiple dimensions, and there are multiple stable equilibria for dividing them.  By exploiting a highly salient cross-cutting issue where each party is internally divided, political entrepreneurs can tip the parties into new configurations.  This is what happened with slavery in the 1850s and civil rights in the 1960s.  This is basically Rand Paul’s gambit for 2016, about which I will write more soon.  A few practical problems with applying this theory to Uber – it’s not salient enough, it’s not cross-cutting, and even if it worked current party leadership would lose their jobs as the party base completely changed.  So a pretty half-assed effort here all around.

Finally, I just want to touch on the consequences for Uber if the GOP’s plan worked, because they are terrible.  If an issue successfully becomes a partisan issue, the natural implication is that the other party lines up against it.  If the GOP succeeded, then Uber would find themselves in the crosshairs of the Democratic Party after the GOP had made the issue highly salient.  Democrats happen to run the cities where Uber operates.  The consequences for their business would be highly negative.

More generally – is it a good idea for minority parties to attempt to activate new partisan issues?  If you believe elections are mostly decided by fundamentals, it is a terrible idea.  By raising the salience of these issues, minority parties make majority party action more likely and in a direction that is likely to differ from the minority party’s preference.


*: Poll was commissioned by Uber.  Most polls on the issue are commissioned either by Uber or the taxi lobby.  This one had the fairest wording – asked whether existing regulations are sufficient or whether it should be regulated more heavily.  

What Should Wages Be? Does the Question Make Sense?

I’m liking the NYT’s new “Upshot” section, which aims to provide a more quantitative and data-informed take on the news (including political scientist Lynn Vavreck!).  There’s an interesting piece from today that could, I think, use a little more detail.  It concerns the decline of truckers – and trucker salaries at the same time.  Trucking companies are having trouble hiring enough talent to meet demand.  The explanation Neil Irwin offers is simply that wages are too low, and that managers are resistant to higher wages that would allow them to adequately expand.  Irwin offers an explanation that is frequently heard, but that doesn’t make much sense – that employers simply refuse to offer a market-clearing price.  Not to beat up on Irwin, because it’s a common argument, but it doesn’t seem to make much sense.

The idea that wages in a free market are “too low” doesn’t seem to hold up to more than casual scrutiny.  Keynesians often refer to “nominal wage rigidity”, wherein managers are unwilling to cut wages in recessions.  It does seem to be real, but this specifically is downwards nominal wage rigidity and is a function of existing contracts with workers.  What Irwin’s argument (and others like it) posit is widespread upwards wage rigidity, wherein employers have a firm anchor to the prices they’re willing to offer new workers.  The anchor comes from…well, it’s not clear where the anchor comes from in this account.  Wage caps can come from heavy regulations or from cartels, but neither applies in this situation.  There are many small trucking firms, and if talent is systematically underpriced they would leap in and seize the opportunity and the supposed wage gap would go away.  Irwin’s explanation implicitly relies on the (appealing) intuition that we know what the market-clearing wage ought to be, but in fact we know no such thing.

If explaining a phenomenon requires us to completely discard all of our Econ 101 assumptions…the explanation is more likely wrong than our assumptions.

World War I Never Really Ended

The world is terrible today.  The US has just begun bombing in Iraq, the Russians are still menacing Ukraine, Syria remains in shambles, the African ebola outbreak continues to worsen, the Israel-Hamas ceasefire has collapsed, and to top it all off Azerbaijan seems to have decided the time is right to declare war on Armenia in a rerun of one of the deadliest conflicts of the Soviet collapse.  This may just be a bluff, it may be serious, or it might be both.  Not coincidentally, last week we just commemorated the 100th anniversary of the beginning of the First World War.  The amazing thing is that (except for the ebola outbreak) all of the above crises can be traced back to the events of a hundred years ago.

The wreckage of World War I still marks these conflicts. The fall of the Russian empire set off vicious wars amongst the various nationalities – including the Azeri-Armenian war and the failed Ukrainian War of Independence.  The Soviet takeover froze those conflicts, but the fall of the USSR booted them right back up and they continue to play out. The Israeli-Palestinian conflict is a direct consequence of the British seizing Ottoman Palestine and opening it up to Jewish settlement.  Iraq and Syria are creations of the war as well, and the weakness of these multi-ethnic, loosely-bound states are the legacy of heir creations as ad-hoc colonial divisions.  In some sense, the blurring of the border under ISIS control may be more stable than the rigid state boundaries that existed before.

World War I deeply broke the prevailing international system, and a century later the consequences are still playing themselves out at the cost of many, many lives.  Something that you should keep in mind when anyone confidently predicts the outcomes of military action.

How to Study for “Jeopardy!”: A Data-Driven Approach

Some kind internet-dweller has put together a dataset of all the Jeopardy! questions up through 2004, roughly 200,000 of them.  This seems like it could be an interesting dataset.  I wanted to ask and answer a specific question: where should a prospective Jeopardy! player concentrate their effort?  The data comes in JSON format and has the following information:

  • 'category' : the question category, e.g. "HISTORY"
  • 'value' : $ value of the question as string, e.g. "$200"
    • - Note: This is "None" for Final Jeopardy! and Tiebreaker questions
  • 'question' : text of question - Note: This sometimes contains hyperlinks and other things messy text such as when there's a picture or video question
  • 'answer' : text of answer - 'round' : one of "Jeopardy!","Double Jeopardy!","Final Jeopardy!" or "Tiebreaker"
    • - Note: Tiebreaker questions do happen but they're very rare (like once every 20 years) - 'show_number' : string of show number, e.g '4680'
  • 'air_date' : the show air date in format YYYY-MM-DD

So after fetching and transforming the data, an obvious first question is this: what are the top Jeopardy categories, and is it possible to game the system?  To my surprise, the show does not have many ‘favorite’ categories: here are the top 5:

  1. Before & After: 587 questions
  2. Science: 519 questions
  3. Literature: 496 questions
  4. American History: 418 questions
  5. Potpourri: 401 questions

This seems like a lot of questions for a few categories, but in fact over this span, there are no fewer than 27,995 categories and 216,930 questions.The top 5 categories represent a whopping 1% of all questions asked, which is…not particularly helpful in targeting your studying. The top 100 categories contain only 11% of all questions.  You might suspect that Jeopardy! has a “long tail” of question categories, and you’d be right.  The distribution of question frequency is on the left, with a logged version on the right.  The long flat line represents categories with only 5 questions; that is, categories that have only appeared once.  There is a very long tail of Jeopardy! categories – mostly because, as long-time viewers will know, often categories have specific and descriptive titles even if their “real category” falls into science, literature, or history.  The punny ones don’t make it any easier, either.



Unscaled on left, log on right for clarity.

Incidentally, the data-generating process isn’t easy to model.  The weird baseline at five combined with a few ones (i.e., Final Jeopardy! questions) resists simple quantification and most common probability distributions fail. I had the best fit by far with a log-normal distribution, but that fails to capture both the extremity of the left tail and the low level of the right tail.  It would be neat if Jeopardy! question distribution mimicked naturally-occurring probability distributions, but sadly this does not seem to be the case.


Oooh, pretty lines.

This whole process has given me a lot of sympathy for IBM’s Jeopardy!-playing robot Watson, because this data is extremely messy.  The long tail of the category labels mean the dataset is not easy to work with at all, and brute-force attacks on the dataset will yield almost nothing of value.  The only way they could produce such impressive performance was with a very high degree of sophistication in natural language processing.  And unfortunately, the data doesn’t suggest that studying Jeopardy! is particularly easy to hack.  There are no real shortcuts in the question distribution or any categories that are that disproportionately helpful.  A survey of the top categories suggests that you should know your history, but “know everything about history” isn’t exactly easy actionable advice.  If you were hoping for a clear and easy answer, I have bad news for you – you can study your ass off and know everything and win the hard way, or build practical weak AI and win the much much harder way.


The Risks of a Russian “Gamble for Resurrection”

Dan Drezner nails it – Putin is backed into a corner, and that should frighten us.  Putin is actually paying massive costs for his strategy in Ukraine, for no discernible gain.  Sanctions are increasingly biting, and with the MH17 incident mobilizing Europe this will only get worse in the near-term future.  And so far, his gains from this strategy amount to: Crimea.  At tremendous economic and financial cost, and for the loss of his client state in Kiev. Even worse, the sanctions are beginning to splinter his domestic base.

One classic IR idea is the “gamble for resurrection“.  This is the idea that leaders, especially in authoritarian states, cannot afford to lose in major crises or conflicts – the regime’s support is shallow, and they might lose control if they show weakness.  So rather than back down, often leaders will escalate conflicts because it serves their best interests rather than those of the state.  The more sanctions undermine Putin’s elite coalition, and the less he has to show for it, the greater the risk that he will decide he cannot afford to lose and will start being riskier and more aggressive.

I’d add one more point to Drezner’s – that sanctions aimed at splitting Putin’s domestic base might well be read by Putin as an attempt to force regime change.  Putin surely understands the above logic, but might well put a different spin on things – that the point of these sanctions are to undermine his base of power and hope for a new government.  If so, the incentive for him is clear – each new round of sanctions must be met with more escalation, because if he backs down the personal costs could be immense.


Get every new post delivered to your Inbox.

Join 119 other followers