A different kind of political geography

In the wake of the surprising (to say the least) electoral result I initiated a few projects to try and understand the politics of the country. One thing I wanted to understand was the impact of demographic and underlying situational variables (e.g. health, income, unemployment, etc.) on how people voted. Was the vote about Obamacare? Was it about lost jobs? Was it all education levels? Or was it all racism? Theories have been floated but I haven’t seen a rigorous evaluation of these hypotheses. What’s below is just an exploratory analysis, but the data does point in some interesting directions.

What follows below are a series of visualizations of a large, aggregated dataset of both demographic, situational, and electoral data. Sources for the demographic and situational data are listed here and the electoral data is from the New York Times.

The type of visualization is called a self organized map. Roughly speaking, each hexagon is a group of counties; the map arranges the counties such that similar ones are closer to each other on the map and dissimilar ones further apart:

hs_diploma

For any given variable (here – the proportion of residents in the counties that graduated from high school) the map is a heatmap. Redder colors means the counties index higher, bluer means they index lower. Here, the upper right are the counties where fewer people have a high school diploma, and lower left are the most educated.

All of the maps shown below are available at this interactive site. (A very similar set of maps, but for voting swings as opposed to voting share, is available here)

Below, we look at the voting share for Hillary. The counties are arranged in the same way as above, but since we’re looking at different variable the map is colored differently. (Confusingly, more votes for HRC are red as opposed to the customary blue for liberals, but work with me here). The reason this map is more organized than the rest is that I used this variable to “supervise” the organization (don’t worry about the details of this – basically it just guaranteed that this particular coloring, which is the reference point, would be organized.)

base_plot

Now that we have the basics in place, we can look at other variables: let’s check a few variables and see if they line up w/ the HRC voting share map. What we can do is draw a boundary around the areas that went strongly for Trump and for HRC like so:

base_plot_w_annotation

And we’ll keep these annotations throughout.

Health and insurance:

The breakdowns for uninsurance and health variables like obesity and diabetes don’t break down along electoral lines: the split goes in the opposite direction, with the highest uninsurance and low health areas going to both candidates:

uninsured

adult_obesity

diabetes

Economic variables:

These graphs should put the “economic anxiety” argument to rest, as the areas with highest unemployment went strongest to HRC and those with the least went strongest to Trump.

unemploymen

earnings

Ethnic Variables:

A few graphs line up pretty well: whiteness and ethnic homogeneity. And whiteness and ethnic homogeneity line up basically on top of each other. This would support the hypothesis that the election for Trump was mainly a cultural (and not a policy) event; white enclaves are reacting against a diminishing place in the cultural landscape – hence the making things great again:

whiteness

homogeneity

See for yourself: 

My code is available here (it is not very well commented or formatted, but it’s there).

As mentioned above, all the maps above are available at this site. If you want to see something similar but with voting swings – the amount the county changed their vote from ’12 to ’16, you can see that here.

 

 

Inauguration Day Fundraiser

I sent this email to a group of my friends that live in NYC. If you live in the area and want to help organize, let me know at thomas.vladeck@gmail.com

Friends,

If you’re like me you’re shocked about what happened on election day. I can’t believe our country decided to put that orange-skinned, pussy-grabbing, bankrupt con artist that’s likely an agent of the Russian government into the Oval Office.

But here we are.

As bad as it is now, he’s not in office yet. And as unpredictable as he is, we have no idea what to expect he’ll do when he gets there (see these two back-to-back tweets as he does a real-time A/B test of governing styles), but it likely won’t be good. One of his top lieutenants already proposed bringing back the House Un-American Activities Committee (a black stain on the history of American Civil Liberties), and if he implements his campaign proposals to remove 11 million immigrants, ban muslims, punish women for having abortions, reinstate waterboarding, and change our libel and slander laws – he’ll be violating the first, fourth, fifth, eighth, and 14th amendments.

Not to mention he’s planning to rip up the watershed Paris Accords (makes sense as Climate Change is a Chinese hoax…) in a year that has seen every indicator of Global Warming reach dizzying new heights.

While we’re on the subject of existential threats, it’s becoming clear where the suspiciously pro-Russia policies of the Trump campaign came from, as Russian diplomats had been in contact with the Trump campaign throughout. This at a time that NATO has put three hundred thousand ground troops on high alert because of a feared confrontation with Russia.

I don’t know about you, but I’m probably going to be a wreck on inauguration day (January 20th). Instead of just watching, let’s do something. And that something we should do is party together, raising money for organizations that are going to fight these abhorrences tooth and nail.

Here’s what I’m thinking:

  • We rent out a space
  • We invite all our friends
  • We have a great time
  • Some great groups get some needed cash to keep the fight going

Who’s in?

Tom

 

Re: the election

Those who can make you believe absurdities, can make you commit atrocities

I owe many people an apology today. I was extremely confident that HRC had this election in the bag and presumptuously advertised this confidence over the past weeks, dismissing any thought that she would lose. It wasn’t an act, and it has made it all the more devastating to me now that she has in fact lost.

Because of my work with Gradient, which does statistical modeling in a business setting, over this cycle a lot of people have asked me for my interpretation of the various forecasting models. When discussing it with them, I was bullish, overconfident, and as it turns out, terribly wrong. I feel terrible for giving people this false impression of security and making the result any more jarring and devastating than it already is.

What was I wrong about? Well, mostly everything – but two very large things stand out: polling bias and the correlation of polling errors. In general there are two types of statistical error: bias and variance. Variance is when you’re dancing around the right result – any one measurement is off but on average the errors cancel out; bias is when the errors don’t cancel. Polling bias (even state polls) in presidential elections has been estimated over many cycles, and has typically been small (about 1%). What polling bias means in concrete terms is that many more people voted for Trump than the polls captured; there was a social movement happening in front of our eyes but that did not show up in the data. I trusted the data and did not foresee the possibility that the polling bias would be so large across the board in swing states.

That leads to the next point. State results are obviously correlated – blue and red states tend to vote together. This means too that polls of states should be correlated. So outcomes and the measurements of those outcomes should be correlated – but should errors be correlated? I had no reason to think so. For example, polls had HRC ahead in Michigan and Wisconsin; I thought that all the correlation between the two would be captured by the correlated polling results across the two states. I did not think that a polling error in one state meant that a polling error in the same direction in the other state was more likely. It is obvious now that this was the case.

This election is going to prompt a rethink across the entire polling and data analysis industry – myself very much included, even though politics is not my professional remit. But that’s small potatoes compared to the vastly more consequential implications of this election: climate change; respect for women, immigrants, and muslims; the supreme court; our security and trade relationships around the world; the nuclear codes, and so on.

As for what happens next, I hope you agree with me that it’s critical that we stay engaged with our country’s politics as opposed to recoiling in horror. Quite literally, I think our country needs us and people like us to fight tooth and nail to preserve our vision of what America is and what makes it great.

Introduction to the Timespace Map of New York

[2017-12-07: This has been updated in a big way.]

This summer I lived in New York for the first time, working for Venmo. Venmo’s offices are in the West Village, and I lived in Park Slope, so every day I would take the F train. As I was getting to know the city I would often look at this map:

 

Looking at that map every day got me thinking that so much of the geography of the city was defined by its public transit infrastructure, not by the lay of the land. What determined if two points were practically far or close depended not on how many miles separated them, or natural features like rivers that may lay between them, but whether or not it was easy to get from one place to the other on public transit.

I started thinking about how we could make a new map of New York, where distances between points on the map would reflect how long it took to get between them, not how far apart they were. It would project time onto the two-dimensional space of the map.

Without knowing how I would do that, I started collecting data. I took the 195 neighborhood tabulation areas of New York, geocoded them into Lat/Long pairs, and started hitting Google’s Directions API for each of the pairs of neighborhoods to see how long the transit route between them took. After doing this over many days in batches (because Google limits how many directions you can ask for any given day), I had a distance matrix that looked like this:

 

Screenshot 2015-09-05 15.20.20

 

With the cells being the time (in seconds) it took to get between the two locations.

Next, I needed to figure out how to lay out these points in a two-dimensional space that reflected these distances in time. Turns out, this is a common problem in data visualization, and the most common tool is called multidimensional scaling. Running cmdscale (in R) on my distance matrix allowed me to lay out the “neighborhoods” of New York onto a 2d map with distances between them reflecting time.

With just this information, about the best I could do was to create a Voronoi diagram of the points scattered about.

 

voronoi

Cool, but we can do better.

I started thinking about these 195 points as a “skeleton” that I could build on. I wanted to project more information onto this map, but I couldn’t just go on adding points to my distance matrix (since the number of entries in the distance matrix goes up with the square of the number of neighborhoods), I had to figure out how to project an arbitrary Lat/Long pair onto this map.

I tried a few things out, but the method I settled on was LOESS regression as it was simple, made few assumptions, and could handle the obvious nonlinearities in the transformation between Lat/Long space and Timespace.

The model was simple (this is the model to predict one axis of the timespace dimension):

model.dim1 <- loess(data = model.data,
formula = dim1 ~ lat + lng,
degree = 2,
span = span)

And the span term allows you to control how much impact the skeleton points have in “pulling” the new Lat/Long points away from their starting point. Here is a gif of different projections of the five boroughs with spans ranging from 0.99 to 0.2: (there a few errors in the borough shapefiles that I’m aware of…)

animation2

So what’s next?

Well, now that I have a method of projecting arbitrary Lat/Long pairs onto the “Timespace” dimension, I am working on doing it with other data sources. I’m starting with subways, but plan to add other features like roads, parks, and the like. Stay tuned!

Variations on a theme

Anyone that knows me has probably noticed that over the past year-plus, I have gotten into photography in a major way. But as much as I post photos, I spend about as much time ogling the photos of more talented and able photographers. One thing I noticed is that the best photographers tend to have a very specific style – for Randy Martin, it’s a perfectly centered subject in a larger scene; for Michael Goldberg, it’s close-ups on the street using flash to illuminate his subject’s imperfections; for Nguan, it’s warmly-lit subjects lost in a moment of ennui.

I guess this shouldn’t be too much of a surprise, as variations on a theme are the rule in other forms of art, like music (bands have a sound that cuts across their songs/albums), movies (wes anderson), paintings (dali), etc. I want each photo to stand on its own and be considered in its own right, and I want the ability to take any photo that I think is compelling for whatever reason. But that’s not how it works; your audience (whoever that is and however many people that is) wants something familiar as well as novel and interesting. Perhaps it’s the commonality between your works establishes a common ground with your audience and without a familiar “vocabulary” they would lack the basis to appreciate the work – but I’m just speculating. 

I don’t (yet) have a style. Or at least not one that I’m aware of. Perhaps it takes time to find what you like. Perhaps it takes time to develop the skills that allow you to impose your style on the scene in front of you. I really don’t know! But it’s on my mind this morning.

Capital in the Twenty First Century

I’ll confess I didn’t finish it (audible tells me I made it about three-quarters through). It is, however, a great book, but it does not make for great listening. This one (and it is the first book I’ve thought this about) really should be read on paper.

The central thesis of the book is that the return on capital generally exceeds the rate of economic growth, and when economic growth is slow (as it has been for much of human history) that this leads to a natural tendency for societies to develop extreme inequality.

I’ve long had a “rich-get-richer” argument playing out in my head, but through a slightly different mechanism, which plays out like this: as people get wealthier, their ability to take risk with their capital increases; as this “risk appetite” increases, they are able to devote a more substantial share of their capital to higher-risk assets with a higher expected return; this, then, would lead to a higher overall return to their investments. Think of a person initially holding only cash, then saving up enough money for a down payment on a home, and then to investing their savings in liquid stocks and bonds. Each class of investment (cash, real estate, equities) is riskier and higher-returning than the last, and this sequence leads to an increasing return on increasing capital. Wealthier people are invested in all these asset classes and more, such as hedge funds or privately-held companies.

So it was no surprise to me that there should be runaway effects to wealth inequality. But the rationale Piketty lays out are quite different. I must confess that I did not completely understand the reasons that returns to capital (as a whole, not the varying allocations of capital that I describe above) ought to generally be at a rate faster than economic growth. But the data show convincingly that it is now, and has usually been so. The result is runaway wealth inequality that has only been tempered throughout human history by capital-destroying calamities, such as world wars, that bring people back to an equal footing.

Where Piketty really succeeds, if you ask me, is in linking this argument to the literature describing the extremely stratified societies of, for example, Victorian England. He saying “watch out, these crazy societies could one day be reborn”. I hope he’s wrong, but he’s given ample reason to fear that he might be right.

Review of Lean In

Recently I finished listening to Lean In, which you should probably Google if you don’t already know it. I listened to it as part of a college-friend-audible-book-listening-club, and kept some thoughts about the book as I listened on. Here they are.

If there is one key theme to the book that deserves singling-out, it is the focus on the individual woman as the unit of analysis. Sandberg’s book, when it’s all boiled down, is a entreaty to women, as individuals, to consider a fuller range of career and family options, and gives many “how-tos” regarding how to accomplish some of the tougher options. She is basically saying “You, promising young woman, here is what you have against you and how you can best navigate it”. So although she does pronounce some big goals, such as 50/50 splits in the executive ranks everywhere, her book is emphatically not about what policy needs to do to change things. This is a pragmatic manifesto which aims to give women more tools to confront, as individuals, the institutional sexism that impedes their career progress.

I’m going to get an MBA quite soon, so I’m a sympathetic audience for career-enhancing tips, and I found much of her advice quite applicable as well as useful. Perhaps sadly, though, much of her advice is either not gender specific or only applicable to elites.

Sandberg’s advice can be broken down into two categories: that aimed at helping women more adeptly confront sexist barriers in their professional lives, and that aimed at helping women to make choices that balance their professional and personal lives. For the former, most of this material is, when one really gets down to it, applicable in equal parts for men For example, when outlining the differing negotiating strategies of men and women, the advice that she gives to her female readers is to use tactics that are, in their essence, a subset of the “principled negotiation” methods outlined in Getting to Yes. Men, too, would be well advised to adopt state-of-the-art bargaining methods, not to mention develop relationships with mentors, and to have some thoughts about their future careers.

Although Sandberg is absolutely right to attack the stigma placed on working mothers from all sides (showing, for example, that children with home care from nannies develop just as well and have just as healthy a relationship with their mothers as those raised by full-time moms. I, for one, am one of these children, so I should hope so.), it is just not the case that many women are in a position to consider many of the options that Sandberg urges her audience to feel better about, such as working more flexible hours or having a nanny. Sandberg acknowledges this, and this does not affect the efficacy of her argument; it does, however, limit its scope.

The most powerful part of the book is Sandberg’s entreaty to women not to put their foot on the brakes of their professional lives until they are indeed confronted with hard tradeoffs. The point of the practical portions of the book being, of course, tools to give women a better set of options in these dilemmas.

The book is littered with references to studies and much of its summarized research is quite compelling. Sandberg erred, though, in making her argument for the motivation of the book; namely, that women’s progress infiltrating all strata of professional ranks had stalled. Her data, which were only long-term statistics, only showed that progress was not yet complete, not that it had slowed down. This mishandling of the data led me to approach the rest of her citations with a more skeptical eye.

It’s an interesting book, but far more so for its cultural cache than the material inside.

What I’m looking for at Wharton

After we shut down Building Hero, my career goals were pretty simple: start and launch new products and businesses. My analysis led me to conclude that this happens in essentially two ways: (a) as an independent entrepreneur and (b) as a product manager (or similar role), usually within tech companies – after all, big companies launch new products all the time. After taking as many coffees as possible, putting out a few feelers, I concluded that (b) was going to be a very tough role to land with my background. And for (a), despite working on Building Hero for two years, I never built up a deep network in tech; I explored an idea for about a month but concluded that I didn’t have either a strong idea to build on nor a network in tech to leverage.

I concluded that business school would be a good chance to learn many of the basics of business-building and the product development process, build a great network alongside a world-class cohort of people, and that my degree would provide me with significant reputational capital. In addition – and I didn’t articulate this beforehand – I think the two years at business school will be a great opportunity to create new things with the resources available to students (not least the amazing cohort of fellow classmates).

On August 4th, I’ll start the first of two years getting an MBA at Wharton. One of the pieces of advice they give you when you’re a matriculating MBA student at a school like Wharton is to take some time to really think about what you want to get out of it, because there will be so many potential demands on your time. So as I’m winding down my current job, I’ve started to think about what I want to get out of the whole experience.

Below, I list out the overview of what I want to achieve, broken down into the following categories: academics, community, leadership, personal branding, and finally, some specific plans.

Academics

As mentioned, academics are one of the four core reasons I decided business school would be a good choice for me. I actually probably weight academics more strongly than most other matriculating students; I’ve experienced firsthand what it’s like to try and build a business without working mental models, and it was a tough experience in which I usually felt lost. Only late in the experience when I started taking classes on product development and reading more widely on the subject did I realize how lost I truly was.

At any rate, a true accounting of my skills would rate me a bit behind the curve in many basic business categories. I’m most excited about classes on product development, entrepreneurial management, marketing, and leadership – as they relate most directly to my career interests – but I’m also excited to get a basic education in finance, operations management, and accounting. So – the basics.

Community

The greatest value in business school is probably in the cohort. It certainly felt that way when I got to meet my future classmates this past Spring at Welcome Weekend. So it would be a crime not to find ways to maximize that experience.

Clubs are a big part of business school – both in terms of the career development process and the social scene. Career-wise, the Entrepreneurship and Technology clubs align most closely with my goals. I am going to check out the Marketing and Design clubs as well, but as long as my career goals don’t change it’s seems relatively certain that the Entrepreneurship and Tech clubs will get the majority of my attention. Socially, the Photography and Climbing clubs map perfectly to my current outside hobby interests, but who knows what will gel.

In addition, I’d like to find ways to engage with both the Wharton and the Philadelphia communities outside of traditional channels. (There is a specific Wharton-led opportunity in this area that I am super-interested in, though.) My roommate and I discussed creating a sort of lifestyle/community email newsletter that we’d curate. We’ll see about that – but I’d like to get involved in something that isn’t just put out on a platter by the school. Which takes me to…

Starting Something (and “Leadership”)

Leadership really is a separate category, containing elements of specific training, academic learning (i.e. the mechanics of emotional contagion), and, I dunno, just having to be a leader. My career ambitions put me at the top of organizational structures by definition (i.e. I’m not looking to be a hugely successful investor), so this is a very important skill to develop.

There are some specific leadership programs offered by Wharton which I’ll get to below, but my number one goal at Wharton is to start somethingRight now I’m agnostic on what that will be; it could be a new club on campus, a civic organization, a company, or even “just” an event. But this is an important goal to me; at the very least I want the practice of creating something out of nothing, and I don’t want to let a prime time and a litany of resources for doing so go to waste.

Personal Brand

Careers are changing – everyone knows that the days of working for years at a single company or organization are gone. As Reid Hoffman puts it:

Whereas we used to have a career ladder, now we have a career jungle gym. Success in a career is no longer a simple ascension on a path of steps. You need to climb sideways and sometimes down; sometimes you need to swing and jump from one set of bars to the next. And, to extend the metaphor, sometimes you need to spring from the jungle gym and establish your own turf somewhere else on the playground.

The value, then, of a personal brand is paramount, as your environment will constantly be changing throughout your career.

I’ve got two primary goals for my personal brand: (a) build a fantastic network while I’m at the school, and (b) to keep writing and use my blog as a tool to promote myself.

My view is that connections are best formed organically; trying to network has never been my thing, but when you actually have things going on – you’re trying to get an event off the ground, you’re building a company, you’re writing a blog post, etc. – the people that you reach out to to collaborate with often become friends and future collaborators. So I’m not necessarily going to devote time to this specifically, but my hope is that Wharton will provide a fertile environment for my network to grow as a form of “exhaust” from all the other shit I’ll be trying to do.

I realized that I really enjoyed writing, and that the process of writing really helped me distill and structure my thoughts, during the time I was putting together sixteen application essays for business school; these essays dealt with serious topics about my life and future ambitions, and they helped me form a vision for the future. So I love writing on this blog, and at the same time, having a well-articulated point-of-view on relevant topics is a great way to establish credibility with other people. So I’m going to write on this blog more regularly (my goal is to post twice/week) and to promote it more heavily. I have readership and engagement goals, but I won’t share those for now. I may have to establish a more consistent theme, but this is something I’ll worry about later.

Specifics

In terms of specifics, I will almost certainly look to get an internship in tech over next summer, which most likely means I will be getting involved heavily in the tech club early on. My view at the moment is that this kind of internship will provide me with the most relevant training and the right kind of optionality to either start a company or to gun for a product manager/product marketing manager role after school. I imagine I will be competing with a number of students with similar career goals, so it will make differentiating through personal initiatives all-the-more important.

There are many specific programs that Wharton offers; the ones that I’m interested in mainly fall into “leadership development” and “entrepreneurial resources”. On the leadership side, the programs that interest me the most at the moment are: the Venture Fellows program, possibly because I did a 30-day NOLS mountaineering course, which was one of the most amazing and personal-growth-catalyzing experiences I’ve had; the Non-Profit Board leadership program – I work at in the non-profit space now, and recognize the inherent difficulties of running these organizations well, but at the same time am a huge believer that civic involvement is an important part of a well-led life; and their Executive Coaching program: growing up I played baseball for all of my youth, and had great coaches shape my development; I am excited to see how “coaching” – outside perspectives and regular feedback – works outside of the athletic setting. I see no reason why it should be much different.

On the entrepreneurial resources side, Wharton has a business plan competition, and a “venture initiation program”. It seems clear that at the very least for practice that I should enter the first; the latter, which is geared around providing resources to actually start a company, we’ll have to wait and see about.

And just as importantly, some noes. I’m not particularly interested in applying to the Leadership Fellows. I can’t really articulate why, but it doesn’t really interest me off the bat. I’m similarly not interested in being on the Welcome Committee – in this case, although it seems interesting, it just doesn’t seem as interesting as many of the other opportunities available, so I’m going to deprioritize it.

That’s the plan for now! I’m sure it will all change.