Ranked Choice Voting

We’re all familiar with how a traditional election works: each person votes for their favorite candidate. The votes are then added up and whoever has the most votes wins. Simple enough.

But there are also obvious problems with this. Consider a race with with five candidates. Two of the candidates get 15% of the vote each, two others get 20% each, with the final candidate receiving 30%. Under the traditional system (called First Past the Post, or FPTP) the last candidate would win with only 30% of the vote. That doesn’t feel right. Voters who are concerned that voting for their favorite candidate may ‘steal’ votes from the ‘electable’ candidate are familiar with another problem with this system (also known as the ‘spoiler effect’).

Ranked Choice Voting

But there are other methods of voting besides FPTP. One method, known as Ranked Choice Voting (RCV), continues to gain traction. RCV is a method of voting where – instead of simply voting for your top candidate and electing the person who receive the most votes – each voter ranks their choice of candidates. If one candidate receives more than 50% of everyone’s top choice, they win. If not, however, the candidate who got the fewest top-choice votes is eliminated, and the election is re-run.

For example, say Jill, Tom, and Alice are running under RCV. Of people’s top-choice pick, Jill got 40% of the votes, Tom 36%, and Alice 24%. Under traditional voting – what’s known as “First Past the Post” (FPTP), Jill got the most votes so she would win. Under RCV, however, since no one got more than 50%, Alice (the one with the least votes) would be eliminated from consideration and the top choice votes would then be counted again.

To be clear, it’s not the case that Alice’s 25% would be given to Tom or Jill. Instead, you take every ballot that had listed Alice as their top choice, look at who they ranked as their second choice, and count that person as their vote. For example, say two-thirds of the people who ranked Alice first (i.e. 2/3rd of 24%, or 16%) had Jill as their second choice, and one-third (1/3rd of 24%, or 8%) had Tom second. In this case, Jill would have 40% of the vote (from the first time around), plus an additional 16% (from the second time), for a total of 56%. Because Jill now has over 50% of the vote, she wins.

There are many benefits of RCV over the traditional First Past the Post (FPTF) methods. It avoids the spoiler effect and gets closer to ensuring that the candidate that’s chose maximizes the satisfaction of the electorate. Several places (like NYC) already use RCV, and non-partisan groups like Fair Vote are advocating for it. For those looking for social proof, it has been endorsed by many Nobel Prize winners, political thought leaders, political scientists, and, yes, politicians from both parties, including John McCain and Barack Obama.

Benefits and Tradeoffs

I certainly think that Ranked Choice Voting is a better system than First Past the Post. But that doesn’t mean there aren’t downsides. There is a great article that runs through the different voting systems and their pros and cons, and the Fair Vote site makes additional arguments for why RCV is better than the other systems. (I’m personally a fan of what’s called score voting, where each candidate is rated on a scale – similar to Amazon ratings – and the candidate with the highest rating wins. That said, RCV is the alternative system with both the most popular support in the U.S. and the most “real world” use, as its the primary form of voting in Australia, Ireland, New Zealand, and many others.).

If you have a choice to support or petition for Ranked Choice Voting over FPTP, I’d recommend it. No voting system is perfect, yet few procedural changes (perhaps along with moving to open primaries and bipartisan redistricting) have more downstream consequences for our form of government. Choose wisely.

What Causal Inference Can Tell Us About Hiring

One area I’ve gotten interested in lately is causal inference. For those of you not familiar, it’s a methodology that attempts to find and validate cause-effect relationships between variables. The key is that it attempts to do so using data without having to rely on controlled experiments. (For an introduction for the casual reader, I highly recommend Judea Pearl’s book The Book of Why.)

One concept I found interesting was the implications of something called a collider. A collider is a variable that is the effect of two or more variables. As a simple example, consider the following:

The way to read this diagram is that fame is a function (or effect) of money, talent, and looks. In other words, fame = f(money, talent, looks). In this example, fame is a collider relative to money, looks, and talent because they all have arrows pointing into fame.

The interesting implication from the book is the following: given that you hold the level of a collider constant, the other variables become dependent upon each other even though there is no causal influence on them.

To understand this better, let’s use an even simpler example: X + Y = Z. In this case, Z is a function of X and Y (i.e. Z = f(X,Y)), so Z is a collider with respect to X and Y:

Here’s the key point: if we fix the value of Z at some specific value (say 10, so we’re left with the relationship X + Y = 10), then X and Y become correlated. In other words, if I know the value of X (say 8), then I can infer Y (i.e. 2).

The interesting finding from causal inference is that this dynamic generalizes. Said another way, for a given level of Z, information about X automatically gives me some information about Y, even if I can’t observe Y directly.

Almost Famous

Why is this interesting? Let’s go back to our fame example. Assuming our causal model is valid, then we can say that for a given level level of fame, if we know something about their level of wealth, we can infer something about their level of looks and talent. If we simplified it down for a minute to say just include looks and talent, then we could say – for a given level of fame – we’d expect that a person who is more attractive is likely to be less talented. (Another way to think about this is if they were both attractive and talented, they’d be even more famous).

I haven’t done an analysis to verify this yet, but it’d be interesting to run an experiment. For example, look on social media for actors who have a similar level of followers (as a proxy for fame). Within that cohort, if the model is valid then you would see a spectrum ranging from the good-looking-but-hacky to the talented-but-ugly.

Counterintuitive Hiring

This finding has interesting implications in many places. Take hiring, for example. Consider, for example, a hypothesis that seniority_level = f(skill, likability). If you think both skill and likability are positively correlated to seniority level, then – for a given level of seniority – consider that the most skilled person is likely to be the one you personally like the least.

These are of course toy examples; the causal structure of real life is likely to be much more complex. But they illustrate both the power of causal analysis and the sometimes counterintuitive truths behind the way the world works.

Process Over Outcome

Consider a group that is attempting to create organizational or political change. I typically see the leaders of these groups take one of two approaches.

The first is to get a bunch of like minded people together in an attempt to “make things happen”. These types of groups mostly agree on what needs to be done. But this approach often suffers from two problems.

First, there are other groups of stakeholders who may not agree that the proposed solution is the right one. Since most situations like this require buy-in from many different stakeholders, this lack of buy-in dooms the solution to failure. Second, even when the group does have the ability to “ram through” their solution without securing broad support, their solution is often missing key perspectives or makes invalid assumptions, which render their solution ineffective.

The second approach is to bring a diverse set of stakeholders together and attempt to get alignment. This approach addresses the problems above but often runs into its own challenges. In many cases each faction has their own favored solution that they “know” is the right one; each group is just waiting to take the floor so they can explain why their solution is the one that will work. Groups may question the motives of the other groups and trust can often be strained.

The Power of Process

The power of the scientific method is that – at least nominally – scientists are not wedded to a specific “answer” but instead are aligned by their agreement on a fundamental process: they agree on a set of methods and processes for determining what is true.

It seems like this would be a better approach to change management as well. Rather than starting by trying to agree on a solution, start by getting agreement on the process by which we will arrive at a solution.

Note that in some cases this approach might mean using a different set of criteria for who is invited into the group. Perhaps it might mean splitting folks into two different groups: a content group who are experts in a given area or representatives of a given constituency, and a process group whose job it is to take the information provided by the content group and use it to inform their process.

Getting key stakeholders in each group to agree to a process achieves two things. It negates any specific emotions or biases that the group might have for or against a given outcome. And it creates psychological commitment to a process, making it more likely that they will agree to the results of the process, regardless of outcome.

This approach does have downsides. First, it requires that each party is willing to give up its “known” solution. It would be naive to think think that ego and self-interest aren’t often at play in these situations. But through a combination of reasonableness and peer-pressure (who wouldn’t want to follow a logical process to get to an answer? *wink*), perhaps these dynamics can be mitigated. Second, because the parties must first agree on a process rather than jumping straight into the content, this approach can seemingly take longer. Of course the only time measure that matters is the time to achieve the desired outcome. And I’d argue that this approach will lead to better outcomes more quickly.

Getting alignment is never easy. Nor is designing solutions that work. But it seems like focusing on process over outcome might help.

Which Way To Miss?

Policy, as in many areas of life, is about tradeoffs. To take a simple example, consider arguments that some conservatives and progressives might make regarding welfare. I’ve heard friends that lean progressive say things like “How can we let someone who is really trying and down on their luck go hungry? We need to increase the availability of SNAP [food stamps].” On the other side, I’ve heard friends that lean conservative say some version of “I’ve seen people who get food stamps just waste them on things like cookies, cake, soda and chips – we need to reduce their use.” Who is right?

The answer, of course, is that they both are. People come in all shapes and sizes. They also vary in their behaviors, values, and ethics. This is what makes policy so difficult: you have one policy, but how people behave in response to that policy can vary widely.

There are various ways to deal with this. One is to refine the policy. For example, current SNAP policy does not allow folks to buy alcoholic beverages with those funds. This can work well when there is fairly broad agreement that such a refinement makes sense. But this can easily end up getting very complicated as you attempt to refine further and further until you end up with a complex mess that is difficult for the consumer to understand and for the regulator to enforce, and where the interaction effects between the various rules can cause unintended outcomes. (Tax policy, anyone?)

Error Types

Beyond some basic “common sense” refinements, however, a better approach is simply to acknowledge that any policy is going to have some “error” in it, and to ask which type of error is more acceptable, and how much? This is basically the same thing as thinking about Type I and Type II model errors in hypothesis testing.

Using the example above, would you rather someone get food stamps that didn’t really need them or not give someone food stamps that really did need them? To be clear, not everyone may agree on the answer to this question, but at least we’re now starting to have a real conversation.

Let’s say you think that it’s better to err on the side of being generous, even if it means some abuse of your generosity will happen. What ratio are you wiling to accept? For example, if there’s one abuse for every 10,000 people you truly help, that seems reasonable. What if it’s 5 people helped for every 1 abuse? 1 to 1? What if it’s 5 abuses for every 1 person truly helped? What if it’s 10,000?

Standards of Proof

Some parts of our legal system are already explicitly like this (or at least try to be). In the justice system there are known various ‘standards of proof’ that are required depending upon what’s going on. For example, a police officer is required to have a ‘reasonable suspicion’ before stopping and questioning an individual. A ‘probable cause’ is required to issue a search warrant or arrest someone. A ‘preponderance of evidence’ or ‘clear and convincing evidence’ is required in civil court (and sometimes in criminal). And ‘proof beyond a reasonable doubt’ is the standard required for a criminal charge.

Source: DefenseWiki

By placing such a high bar for evidence, we as a society have made the choice that we would rather let a guilty party go free than convict an innocent one. According to Wikipedia, it is estimated that between 2.3 and 5 percent of all U.S. prisoners are innocent. Is that an acceptable error rate? That’s an open question, but at least its a tractable one.

I’m not saying that the details of individual policies don’t matter. Clearly they do. And of course there are other real considerations, such as cost. But when there is disagreement it may help to start the conversation by asking “which type of error are we more willing to make?” and “by how much”?

Systems of Poverty

Several years ago, I participated in the AdvancingCities Initiative run by JP Morgan Chase – cities and/or regional groups were to propose projects that addressed one or more of several key ‘focus areas’ for the initiative, all generally based around improving the major cities within the region. As I am based in St. Louis, I got involved there.

E2E: Education to Employment

My particular proposal was the development of an ISA-financed “Education to Employment” pathway where various stakeholders (i.e. government, employers, schools and non-profits) would coordinate their activities to provide educational and other support services needed to help folks from low-income backgrounds get living-wage jobs. These investments would be financed by an ISA, where the student would start paying back a percentage of their income once they started earning over a threshold amount. The proceeds from this payment would then be divided across the various organizations within the ecosystem according to an agreed upon formula.

It was my hope that this would provide an economically sustainable way to continually reinvest in our communities.

Mapping the System

My proposal wasn’t chosen and we ended up going in a different direction. But in the course of those discussions I learned a lot about the causes of poverty and the feedback loops that often make it very difficult both to escape individually and/or address systemically.

When faced with complex problems like this, one of the things I like to do for my own sake is to develop what I call “causal maps”: basically simple digrams that map out the cause-effect dynamics at play.

The way to read the maps are simple – it’s mainly [Cause] –> [Effect]. I find that doing this often helps me distill a lot of complexity down into something that I can understand, while also highlighting the interconnected nature of the problem.

Draft effect map of poverty
Draft Cause-effect map of poverty

Above is an example of a simple one I created to understand the dynamics of poverty. Now I certainly don’t claim that this is complete or that I am any type of expert. I am mainly sharing in case others find it interesting and to encourage trying this mapping approach and see if it works for them.

The A/B Test for Decision Making

Confirmation bias – i.e. overweighting evidence that supports your view and underweighting evidence that doesn’t – is a well-established cognitive bias. Indeed, it often seems to be the case that when people are presented with evidence against their beliefs, they simply retrench further.

Several years ago, I came up with a trick that helps me avoid this.

The method is simple. In my mind, I make two columns, each of which represents one ‘hypothesis’.

For example, a few years ago my wife made the claim that I was a “bad driver”. I of course immediately became defensive and thought of all the evidence I could that supported my being a good driver: I had never been in an accident; had only received two speeding tickets since I started driving; etc.

Using this technique, I picture something like this in my mind:

For each piece of evidence, I then ask which hypothesis that evidence supports better and put it in that column. For example, let’s say we had the following pieces of evidence:

  • In my driving career:
    • I had gotten in one accident
    • I had gotten two speeding tickets
    • I often parked terribly
    • I had been in many near accidents
    • I seemed to drive much worse when other people were in the car
    • When asked, other people rated me as ‘below average’.

Using my method, in my mind I saw something like this:

Based upon this, I had to concede that the balance of the evidence suggested that, while I may not be a terrible driver, I certainly wasn’t as good a driver as I had thought.

I think the reason this framework works (at least for me) is that it forces me to start by treating the probability of each hypothesis being right as equal and then to consider all the evidence and how each piece supports the hypotheses.

I use this general construct all the time. Is a person being malicious or just lazy? Does God exist or not? Is a given policy likely to be helpful or hurtful?

Now if I were doing this more rigorously I suppose I would instead ask whether each peice of evidence refutes or falsifies each hypothesis. I’ll work on that. In the meantime, I’ve found this practice to be easy to do and effective in helping me think more clearly.