On Big Data and the Human Spirit

Big Data is more human than traditional data. This is in part because traditional data, in terms of scale, is so much more directly manageable. It’s still pretty hands on, especially in the social sciences. This means that over time we’ve gotten very good at working the human element out, or at least minimizing it. Big Data, not necessarily because of its size, but because of its indirectness means we have a great deal less control over where and how it’s collected. Moreover, it’s often collected as the electronic traces humans leave behind as they go about their lives. That’s all they typically are traces—not a data point in the traditional sense. When you conduct your own survey or download the General Social Survey, each data point is an answer to a question. You know what the data point represents. This isn’t so with Big Data because all your data points really are is a clue. And in order to know what it means, you must exercise that quintessential human faculty of sensemaking. Accordingly, modes of sensemaking are as diverse as we are, each with our own unique brand of it rooted in our upbringing, education, experience, chance, and perhaps even our genes. There is a growing library of techniques, or algorithms, available that may assist the analyst in generating sense from Big Data. That means use routines in order to sift through the universe of clues to find the patterns that tie them together into meaning. But while computational and likely embedded math and logic, they are not mathematics or logic. Those are closed systems of self-referencing definitions, defined by their inventors as perfect. The Pythagorean Theorem only works perfectly with imaginary triangles because no perfect triangles are known to exist. On the contrary, algorithms are merely the computational expressions of someone’s own sensemaking faculties—an externalized digitalization of a human mind. When you begin to create your own algorithms, you’ll be uploading a small bit of yourself to carry on your own human uniqueness inside a macroscopic, but highly simplified, virtualization of the real world.

As the science of Big Data matures, I expect to see a resurgence of the humanities and psychological sciences. Big Data’s promise, of course, is that it stands to give us a greater lens on humanity than ever before. But in fulfilling this promise, we will necessary reach advancements in our understanding of the human being as a thinking and feeling being, which may in the end, allow us to transcend entirely the limitations we’ve forever imagined ourselves to be bound.

The problem with polarization

The lasting hangover following this binge of politically partisan fervor will be that even moderates regard each other as partisans. Tragically, moderates were the ones who had a chance of finding common ground and making incremental progress. 

McCutcheon v. FEC is a blow to American democracy.

In the McCutcheon v. FEC decision (download here) the Supreme Court argues that people would be "delighted to see fewer television commercials touting a candidate's accomplishments or disparaging an opponent's character." The decision continues "[m]oney in politics may at times seem repugnant," tolerating speech we don't like is just the price we have to pay for our First Amendment rights. This seems to reflect a fundamental misunderstanding of what people are concerned about. No, nobody likes the endless campaign ads, but we're not upset because they're interrupting Dancing with the Stars. The proper conception of the problem that respectfully acknowledges the concerns of ordinary citizens is that unlimited campaign donations as free speech implies that people with more money get a bigger voice. We cannot ignore the embedded moral hazard, i.e,. the perverse incentives that encourage politicians to screen out the masses in favor of a handful of really, really big donor fish. Rather, the Supreme Court argues talk of corruption must be constrained to clear cases of quid pro quo, dollars-for-influence transactions. I'm sure these Justices are not so naive that they don't realize such claims, like perjury, are virtually impossible to prove.

What's most worrisome is how accepting we, collectively, have become of this practice. Enshrined in the court's decision is the argument that the current aggregate limit on campaign contributions from individuals to candidates ($48,600) allows individuals to contribute the per-candidate maximum of $5,200 to a mere nine candidates! According to data acquired by OpenSecrets (cited at FiveThirtyEight), this deeply unfair limitation on free speech applies to fewer than 600 people. Setting all the high-sounding legal philosophical debate aside for a moment, why are the Supreme Court's priorities such that they are willing to gamble with the liberty of hundreds of millions, for fear that one of these precious few 600 should be denied the opportunity to buy themselves entire legislatures? I argue that this is because this kind of corruption goes well beyond quid pro quo. In politics--as in most everything--priorities follow money. Money has the power to set priorities, by "taking something all the way to the Supreme Court!" Could you do that if you felt your rights weren't respected? No, because you don't have money, and as long as money is equated with free speech nobody will have time to listen to you. Competition ensures that those who attempt to defy this rule lose out to those who heed it. Natural selection.

This 5-4 decision (along party lines) is yet another blow to the integrity of American democracy. It's a hack from the most undemocratic branch of government (by design) that shifts the allegiances from the (by design) most democratic branch of government away from their popular constituencies to a small club of landed patricians.

Bill Nye vs. Ken Ham follow-up thoughts.

I have to admit after watching the debate, I now agree with Richard Dawkins that we shouldn't debate with creationists at all. Nye did well, but imho it wasn't an event he could "win". Ken Ham peddles young earth nonsense for a living--and he's very good at what he does. He's not bound by what science actually says or does and is at liberty to use virtually rhetorical device no matter how fallacious or at odds with reality. He's extremely sophisticated. His arguments and claims are carefully constructed to possess a surface of validity for viewers with a basic, high-school level science education--just enough to give the illusion of scientific credibility to the untrained eye. The end result, I fear, was that a large number of viewers were left with the impression that there really is a scientific debate about whether or not the Earth is 6000 years old. This is a terrible loss because it not only gives radical creationists intellectual cover for their mission to inject religion into science classes, but may find traction among those who were uncertain and/or motivated to believe it. This, I think, is not a trivial fraction of America.

A Partial Explanation for the Shutdown and Thoughts on Outcomes.

Texas congressional districts around Austin

Hello, I know I haven't been keeping up with my blog lately, but sometimes thoughts are a little hard to get together on Facebook post. Many of you are struggling to understand why this current crisis of competent governance is taking place. We're all struggling, including us in the political science world, because the reasons are numerous and difficult to disentangle from one another. In this post I'm just going to describe something which I believe is a key factor. This is a quick and dirty analysis and my political scientists will probably skewer me for not going into district level statistics, but this is for normal people.

Look carefully at the map above. That's a congressional district map of the city and surrounding areas of my current residence Austin, TX. It's tree lined streets are home to a diverse and relatively urbane population famously regarded as 'weird' by other Texans and 'normal' by the inhabits of the populous coastal cities like New York and the San Fransisco Bay Area. While I think this can be overstated, the presence of the massive University of Texas with some 50,000 undergraduates alone and a thriving high-tech sector does give this city-sized town a unique character.

So this morning I decided, "Hey, I don't really think this is going to do anything, but I'm going to call my congressmen and give him a piece of my mind". I got online, punched in my zip code was surprised to find that my representative in this little hipster hamlet part of north Austin I find myself in these days is represented by a Republican. Well now that's surprising, right? Not really, because in 2003 the Republican controlled Texas legislature (lower house) passed a controversial redistricting law designed to ensure Republican majorities. In 2004, the Texas delegation to the US House of Representative went from 17-15 in favor of Democrats to 21-11 in favor of Republicans, a ratio of about 2:1. The ratio of people voting Republican to Democrat was about 61:38, or about 3:2. In a perfect world you'd still expect to see a Republican majority, but not to this extreme. Despite a reapportionment based on the 2010 census that added 4 more districts, the breakdown is still 24-12 (2:1) in favor of Republicans. So what explains the difference? This is why gerrymandering is a big deal.

Let's look at this simple example from Wikipedia:

Gerrymandering example

Okay, examine these squares with the circles in them. Now imagine that 16 of those circles represents the number of people who are supposed to be represented with their own district, and their colors correspond to their party identification, say, Tweedledees (green) and the Tweedledums (pink). Now Senate districts, or states, are always the same. They don't change. But congressional districts must be drawn. In this little example, there are enough people to place them in 4 full districts. How do you draw the lines? Well most people would say something like the box on the left. It has a nice symmetry, right? It's easy for people to know what district they're in, they're nice and concentrated in one geographic area, and they contain about equal proportions of Tweedledees and Tweedledums. Democracy feels good, bro. Each party should have a roughly equal chance to win the district and will have to their best to gain the winning vote with, I don't know, meritorious arguments perhaps. Unfortunately, that is rarely how things happen. Districting is typically done by legislatures, which already have a partisan composition. So they're going to draw with a purpose: to win as many districts as you can.

Examining the box on the left, we still count exactly four districts drawn over the exact same population. Keep in mind that in "single-member districts", if you lose a district it doesn't matter by how much. It's "winner-take-all". So what if the Tweedledums, who happened to be in power and doing the redistricting sayd, "Hey, why don't we totally draw all the Tweedledees into a single district. They'll win by a landslide, but then there won't be enough of them to beat us in any of the other districts! That way, instead of the usual 2-2 breakdown, we can have a 3-1 advantage like forever and ever!" Now if you count out the votes in each of these districts, you'll find there are more pinks in all districts except 1. Same population, same beliefs, different electoral outcomes.

This is exactly what happened in Austin and many other cities around the Texas and the nation. Of all those 7 districts intersecting this liberal bastion deep in the heart of Texas, 2 are Democrat. Over the past couple of decades, Republicans have been on a redistricting rampage. Despite solid Democratic majorities in the Senate and two Democratic presidential victories, the House remains safely Republican and probably will for the foreseeable future regardless of even moderate-sized changes in how people feel about them.

Well, sadly for the Republicans there have been some Unintended Consequences of their redistricting successes, which are partially to blame for our present crisis. You see, many Republican districts are so unshakably Republican now that there is virtually no chance a Democrat could ever win. The only sense that theses seats can be described as "contestable" is not during the general election, but during the primary. The primary is when parties vote to choose who their candidate for the seat is going to be, not who the seat will go to. In other words, Republicans are only vulnerable to other Republicans.

Now do you recall that "single-member district" and "winner-take-all" stuff I was talking about earlier? Well, the explanation is a bit more technical than I want to go into here, but there is this thing called "Duverger's Law" which is like one of the closest things we have to a scientific "Law" in political science. Duverger's Law says that single-member district, winner-take-all electoral systems inevitably lead to a two-party system. (Sidenote: this is why we have two major parties in the US, not three or some other number as they do in other countries). In many Republican districts, the Democrats are effectively non-existent. Well, one thing Republicans can't seem to grasp these days is that A LAW IS A LAW.

Per Duverger's Law, these Republican districts are essentially splitting into their own two-party systems, where the Republicans are now with duking it out with an insurgent Tea Party. Though still nominally a "faction" within the Republican Party, the vacuum of Democrats has created space for the Tea Party to break off on it's own and challenge traditional Republicans in the primary elections for the right to run against a straw Democrat for the House seat.

This is what Boehner, as a leader of what remains of the traditional Republicans, is so afraid of. His party--or what is now more aptly described as his faction of the party--is getting eaten alive from within as a result of over-gerrymandering. The 100 or so of his remaining backers aren't facing serious Democratic challengers who will force them toward the center, who will compel them to have a bit of Obama tolerance, or at least not punish them for any semblance of willingness to negotiate with him. On the contrary, Boehner's Republicans are scarred out of their wits of extremely well-financed (that's another story) Tea Party challengers who will pounce on any perceived willingness to negotiate with Obama and trounce them among the Obama-hating hardliners voting in the primaries.

As a result, we now have this bizarre situation where the Tea Party has tactically put their boot over Boehner's throat by sticking a "defund Obamacare" rider on the ordinarily routine continuing resolution, or CR, that allows the government to spend money and stay open. If Boehner blinks, "his" party (and his job) are done in 2014. If he holds fast, all the "non-essential" activities such as the EPA, NASA, Department of Education, and the FDA that Tea Partiers hate will remain closed. Either way, the Tea Party wins.

The only way the Tea Party could win even more is if Obama and the Senate Democrats agree to pass one of these "mini-CRs" that only piecemeal fund the government programs conservatives like, such as veteran's benefits. If the Democrats allow that to happen, the Tea Party will pretty much have the government they want and will never agree to anything again. Tea Party wins.

How do we get out of this? Well, Obama, the Dems, and the old school Republican Party of Reagan, Dole, and Boehner don't seem to have any good plays. For Obama, he's just going to go on television and ask very nicely and sincerely that Boehner call the vote (watch video here. Watch it. Srsly.). But between the two of them I think they're just going to let the clock run down on the "Debt Ceiling".

In 8 days the government doesn't just run out of the authority to spend money--it runs out of money. That means the government not only stops functioning, it stops paying its bills. From an economic standpoint, this is far, far worse. This means that the US defaults on its debts and loses credit worthiness. Ask any economist and they will scream at you how (literally) even suggesting the US should default could cause lasting damage to not only the US economy, but the world economy. Again, the explanation here is pretty technical, but anecdotally if the US doesn't pay its debts, then its word is worth nothing, and all the things backed by its word (i.e., the US dollar) are similarly worth nothing. The US Dollar is by far the most dominant "reserve currency" in world. This means that US dollars are what most of the world's governments store their wealth in. It means that the US dollar is a base unit facilitating the exchange of goods and services across borders. I'm not going to go into it further, but suffice it to say that the impact would likely be seismic, and categorically unlike anything any of us have ever experienced. There would be no precedent, as the US has not defaulted on its debt once in it's 237 year history.

This is a figurative "nuclear bomb", in the words of Warren Buffet, and it is set to go off in 8 days. Boehner has his finger over the cancel button, and he's going to see just how long he can let his finger hang there before either Obama or the Tea Party blinks. While the Tea Party may hate the US government, you would assume they don't hate the US (or maybe they do and that's why they've been packing all that gear). If neither of them do, then we'll have to see if he blinks.

The media is not your friend.

“If anything good can come out of such a horrible event as today, maybe it will be a wake-up call that we have enemies that are out to destroy us...They never let up and in many ways they are more treacherous than they were before September 11th because they’ve metastasized, they’ve morphed and many of them are under the radar screen, and that’s why we have to be so aggressive.” -- Peter King on Fox News

Terrorists pose a vastly greater threat to your civil liberties than your person, but in this they work through intermediaries.

On Stupid Questions

Was thinking about all the things we don't know. This has got to be one of the hardest things to do because it's a lot. In fact, just how much is difficult to appreciate because it feels like we know so much. As a practical matter, if we've gone 200 thousand years without a single one if us knowing something, we probably have some really strong assumptions about it. We would feel these as the most obvious things in the universe, so patently true that we wouldn't even think it required explanation. I'm talking bedrock assumptions here, guys. Do you know what this means? It means that some of the stupidest questions could contain the greatest discoveries. Is the Earth flat? Were the animals always like that? What is air made of? Think about this the next time you hear a stupid question. Give it a moment before you blow it off and dismiss it to someone's ignorance, and check your assumptions. Maybe there is a discovery behind one.