Introduction

It is perhaps better to start this article off by clarifying what it is not rather than what it is. First, this is not a comprehensive review of RFK’s policies and what he stands for (there are far better places to seek that information). Second, this is not meant to convince you to vote one way or another based on policy and beliefs (again, there are far better places for that too). So then what the blazes did I write this for? Well, the motivation for this article comes from multiple conversations with friends and family who want to know more about voting for independents in general and RFK in particular. Addressing issues such as,

• “Is it a wasted vote?”
• “Do I vote for RFK to make a point?” / “If we do not vote for Independents, then how will they ever win?”

I believe that these are important questions to ask and I hope to address them in this article. In order to answer those questions I will first explain the voting system in place and various strategies that can be used to win and election,

• Differences between Parliamentary (such as the UK and India) and Winner-Takes-All Democracy (USA).

• Splitting the vote - what it really means. Different kinds of potential RFK voters and why they matter.

• Strategic Misreporting - why people who say they might vote for RFK might not actually vote for RFK but simply want you to vote for him.

As a recovering Game Theorist, I love to look at elections as “games” and therefore I will use the word “strategy” a lot. A strategy in this sense is an action (in this context voting for a candidate). In the game theoretic structure, we assume that a player (i.e. YOU) is playing to win. But what does it mean to win? In this context, winning means getting policies you care about enacted. I will also address a little, the issue of voting “to make a point” about the current system and why I feel like that is a bad idea. But for the most part, I assume that the reader, wants to get policies they care about enacted.

Equally important, I will assume that political parties have atleast some motivation to get elected. While getting elected is not the only motivation of political parties, it is certainly a very important one and allows us to separate out our strategies for voting for them.

Differences in Democracy

Perhaps the least understood part of this discussion is the inherent difference between Parliamentary democracy and Winner-Takes-All Democracy (this is technically called Representative Democracy, but I feel that the term obscures its meaning). Before understanding what you should do, it is perhaps worthwhile to understand what the system you are voting within intended for voters to think about. This could be quite different in both systems and have vastly different implications. Usually, the choice of system has more to do with the history and socio-cultural context at the time of setting up the democracy. It is very difficult to argue (vehemently, at least) for one over the other. But certainly, one should try to understand why a particular system was chosen and at least try to engage with viable strategies within that system.

For much of this article, I will consider hypothetical political parties, the first two are large and usually get most of the vote share, the independent is small. KH, DJT and RFK. KH, DJT and RFK (an independent). I will consider two hypothetical elections, one in a Parliamentary democracy and one in a Winner Takes All democracy.

Parliamentary Democracy

Consider candidates with the following vote shares and a seats in the “Parliament” in a hypothetical parliamentary democracy (number of seats won, in brackets).

• KH : ( seats)

• DJT : ( seats)

• RFK : ( seats)

In a parliamentary democracy, KH narrowly wins the election. However, (and this is a big caveat), every time a decision is needed to be made, any one party would need to form an “alliance” with some or all of the other parties to reach the mark. This means that a significant number of independents need to be swayed in order to pass a law (by either side). By the same token, DJT’s influence is not insignificant as they need to sway just more (than KH) independents to pass laws they want. This system comes with a clear message to the voting population’s strategy, you can (and should, if you want to) vote for a party that is smaller than the other two and their voice will be heard at every vote. This system also comes with a clear disadvantage, you need to appeal to independents at every voting instance. This is particularly worse when you consider a situation like this,

• KH : ( seats)

• DJT : ( seats)

• RFK : ( seats)

In situations like this, RFK can hold up legislation that almost of the country wants. Bear in mind, that bills in any democracy do not work in isolation, so RFK can hold up a super important bill (Free Childcare) that even their want in exchange for a bill that only their want (Bitcoin deregulation). There are two other future implications that are essential to understanding the Parliamentary system.

• The first, is that representative democracies encourage a proliferation of independent parties. They do this to the extent that the word independent party loses all meaning, and there are just a large number of parties that cater to ever more niche demographics that can sometimes seem hilariously contradictory (Pro Environment, Pro Socialism) and (Anti Environment, Pro Socialism).

• The second, is that “winning” in a representative democracy ends up being one of two things. You either get of the seats in parliament or you form a coalition that adds up to using various smaller parties. In such a coalition, parties will often “give up” a few of their essential ideas or concepts (Environment) in exchange for passing laws that support another (perhaps more important) essential idea (Socialism).

Notice, that voting for more and more independent parties does not lead to more diversity in voting ideologies, it just means that the reduction in diversity is left up to the party representative not the voters.

For example, say you voted for a pro-Environment, Pro Socialist party. Since they are a niche party they formed a coalition with a Socialist party and gave up on Environmental regulation. Now had you known the full result of the election in advance, you might not have wanted to give up on Environmentalism, you might have given up on Socialism instead. For instance, you could think, if I cannot live in a cleaner environment I might as well have free markets.

This paints a picture of a democracy that is very unstable. It is. Since the resolution or tolerance between conflicting ideas takes place at the parliament it is very difficult to gauge what issues are deal breakers for the voting population. But over time Parliamentary democracies tend to form major parties with a constellation of smaller parties that reflect minor interest groups. Governments are formed by one of the two major parties and a collection of smaller parties. We now turn to the other case.

To fix the issue of stability and to reduce the outsized influence of smaller parties, another form of democracy has been proposed that addresses these issues directly.

Winner Takes All Democracy

It is a bit complicated to show an exact example of representative democracy in the US, but this example is a pretty good representation. In this example, there is no parliament, there is just a president, who can do whatever they want for the length of their term. Consider the vote share example as before,

• KH : ( seats)

• DJT : ( seats)

• RFK : (11 seats)

In this example KH, can pass all the laws they want. It does not matter that they do not have of the vote share. Notice, also that more people did not want KH to be in power. Potentially all of RFK supporters (more on this later) could have preferred DJT to KH had they known the results of the election before hand.

What are the implications of this kind of democracy?

• First, notice that after the election the elected person is essentially a dictator. There is no need for any negotiation or working with any other parties. This is not a bad thing, since much of the confusion and instability of Parliamentary democracy is done away with.

• Second, notice that there is a strong disincentive for other political parties to form since even at fairly high levels of representation you can end up with seats. Consider this example,

• KH : ( seats)

• DJT : ( seats)

• RFK : ( seats)

While people who voted for KH might definitely consider voting for her again, some of the supporters of RFK might consider either :

• Not voting at all - which is why voter turnout is such an issue in the US elections

• trying to persuade DJT to accept them into their party and fight for change in some of it’s core values (maybe considering the environment more).

Summary of Differences in Democracy Styles

The key takeaway is that in both systems you have to eventually reconcile your differences to reach that mark. In the Parliamentary system you leave it up to the person you vote for, no matter how small their party is. But in the Winner Take All system, you have to do it yourself, or you risk coming away with nothing (hence the Winner Takes ALL!). Again, either way, some (or most) of your ideologies will be resolved to reach a decision.

Opinion : So What Should You Do?

Well, one thing is clear, since the US is a Winner Take All system you should reconcile your differences with the major parties and place your vote there. While it was not clear to me why this system was chosen in the US, it seems that the pressure of reconciling one’s differences is on oneself. This system is perhaps why we have a two party system in the first place. The motivation for a voter to vote for an independent is very low (but there is one situation in which it makes sense, more on that later) to the point that it has prevented the formation of more parties. Which is why it is ironic that many independents run on a ticket of plurality of opinion but do not actually advocate to change the actual voting system so that more political parties are motivated to coalesce around different combinations of ideas. But short of that, it is up to you to vote for a major party after giving up on some of your ideals.

Implications for Reconciling Differences

If you are reading this far it means you are at least considering voting for the major parties. One thing is clear when reconciling your differences, you need to figure out which party you would vote for if your top choice did not exist. Thus two kinds of voters exist,\

Where, means is that if you would vote for over . For instance, if after casting your vote for RFK and seeing he lost you would rather DJT won (had you known RFK would not have won), that means DJT is your second choice. Thus, imagine a world in which RFK lost and think about who you would have preferred. That is who you should vote for. Similarly, if you voted for RFK and DJT won, and you wished that you voted for KH, then your second choice is KH.

There is however, one (and only one) situation in which you should vote for RFK and that is the situation in which you are truly indifferent between DJT and KH. That is,IF, on the day after the election you truly do not care if RFK lost. I think that such candidates are likely to be of two kinds (and I do not think readers of this article are likely to be either).

Non-voters : They would probably have not voted any way. If you are going to vote if RFK was not running then this is NOT you.

Ideologically inconsistent : Since independents and RFK generally seek to appeal to both parties and therefor take centrist positions, it is not possible for someone to be truly indifferent between KH and DJT. For example consider the following policy positions, - RFK (Pro-Life, Pro-Environment) - DJT (Pro-Life, Anti-Environment) - KH (Pro-Choice, Pro-Environment)

If you really are indifferent between KH and DJT then you are indifferent between (Pro-Life, Anti-Environment) and (Pro-Choice, Pro-Environment). This is unlikely, since these are such salient issues, you would certainly have an opinion on which you would rather have. If you really are indifferent about such important issues you are not an ideological voter and are motivated by something other than getting policies you care about enacted. This could be someone who votes for RFK to “make a point” about the current system. But equally this could be someone who votes based on personality rather than someone voting on issues alone.

Strategic Implications

Interestingly, it is in the interest of the party that thinks they will lose to promote the independent candidate. Consider the following strategy by DJT,

• Promote RFK as an independent (ask your donors to donate to him).
• Appear as similar as possible to RFK (public appearances, phone calls etc).
• Make sure that RFK is on the ballot in as many states as possible.

With this strategy it will be possible to make it appear like RFK is very similar to you but different enough from KH thereby ensuring that your vote base is intact but people will defect from KH.

Strategic Misreporting

There is another more complex issue that is known to occur in voting. The best way to understand it is to understand that people voluntarily disclose their voting strategy and that this strategy is never verified. Essentially you can say you are going to vote for any candidate and no one will ever know if you did or not. People misreport for a variety of reasons, including embarrassment, social pressure and privacy. With the rise in far right parties in Europe, people are less likely to admit that they voted for them. However, one of the most interesting reasons to misreport is for strategic reasons. Consider the following strategy,

• You are a DJT voter and you know that RFK is more likely to take votes away from KH than DJT.
• You tell people you are going to vote for RFK, this will encourage other people to vote for RFK.
• This will make it more likely that KH will lose votes to RFK but not DJT.

Thus when discussing your voting strategy it is important to remember that a person whose second choice candidate is KH and whose second choice is DJT are fundamentally different people.

Conclusion

• “Is it a wasted vote?”

Yes it is, for reasons above the American system expects you to reconcile your differences with the major parties and then cast your vote. If not, you will come away with either :

- your third choice candidate winning implementing policies that are objectively worse for you. 
- you vote for an independent but the people telling you to do not (strategic misreporting).

• “Do I vote for RFK to make a point?” / “If I do not then no independent will ever win?”

No you should not. The reason that independents do not win has more to do with the system than the fact that they do not get enough votes. Even if an independent ends up with very very high percentages of vote share they can end up with no representation. The system is inherently Winner-Takes-All, now you could ask, “why not change the system?” and that is a good question. Unfortunately that would need to be done by the major parties and they have no incentive to do so. But guess what, the best way to do that is to vote for a candidate from the major parties who has a policy of changing the voting system. Best of luck with that.

In the past many candidates have been independent and have garnered huge amounts of popular support (at the primary stage), but these candidates have inevitably joined either of the two parties. So what ends up happening is one of two things,

• if the major parties think an independent is popular and risks a big chunk of vote share, they offer them a ticket.
• if the major parties do not view them as a risk they ignore them and hope they do not take too much vote share. If they do take vote share this has the effect of penalizing the candidate who has less fanatical (nationalistic/ personality driven) supporters since they are more open to truly voting based on ideology.

I think that rank order voting is a good system to implement in the US, and advocating directly for that is a better strategy than voting for an independent. As I said, it is funny that independents do not directly advocate for this system, but it is likely that they are not able to get enough votes to be taken seriously.

Let us conclude with an example of rank order voting. In this voting system, instead of voting for candidates you express your preferences for all the candidates. And the candidate with the least points WINs. That is, not only do you care about how many ballots had your name at the top, but also considers how many people had you at the bottom.
KH\succ DJT\succ RFK (1)
KH\succ RFK\succ DJT (40)
DJT\succ KH\succ RFK (1)
DJT\succ RFK\succ KH (36)
RFK\succ KH\succ DJT (15)
RFK\succ DJT\succ KH (7)

KH points : 41 * 1 + 16 * 2 + 43 * 3 = 1 + 32 + 129 = 162
DJT points : 37 * 1 + 8 *2 + 55 * 3 = 37 + 16 + 165 = 218
RFK points : 22 * 1 + 76 * 2 + 2 * 3 = 22 + 152 + 6 = 180

This example proves the benefits of rank order voting since you can notice several things.

• KH wins in both systems, if you have enough first place votes you are the winner pure and simple.
• DJT’s loss was made worse by this system because of the huge number of people who had him at the bottom. This is not surprising for the people who had KH on top of their ballot. But because of the huge number of people who had RFK on the top of their ballot but DJT at the bottom of the ballot.
• RFK is not as bad a candidate as it seems, even though he had only 22 first place votes, when considering his second place votes he is actually not a bad candidate.

In the rank order system you can use your third place vote to essentially veto a bad candidate, it essentially says this is who I prefer at the top (RFK) but I definitely don’t want my 3rd place candidate (DJT) I would rather have (KH). This essentially allows the two different kinds of RFK voters to express both their preferences.

The Bayesian Policy Maker

Ah they say, so here is what you do, you see its very simple. You gather data, you gather all the facts, and then you do the statistics you see, and then you make your decision. You see, a modern policymaker shouldn’t bother with the inconveniences of a ideology and emotions et cetera, that stuff is for amateurs you see.

Also, you attach a token picture of poor people being poor in a 3rd world country on a website for RCTs (the cover image above is taken from one such website, not sure why it is relevant to their study) and you are well on your way to success!

Rarefied air of RCTs

In the hallowed halls of economics, evidence-based policy has long been the order of the day, with statisticians and economists working hand in glove to unravel the mysteries of various policies. This delightful dance of data was often accompanied by the sage nods of experts. Enter the Randomized Controlled Trial (RCT), purportedly requiring nary an assumption nor a whisper of prior knowledge.

Ah, but herein lies the rub! Some social scientists, in their wide-eyed admiration, have crowned RCTs as the veritable holy grail of evidence, declaring that any nugget of knowledge gleaned from an RCT is the unvarnished truth, thus tossing aside the cumbersome baggage of expert opinion. Combine this with the dazzling allure of Bayesian epistemology, and we have a recipe for an unearned swagger in the land of causal inference.

This article aims to lift the veil, to show that the RCT, for all its bravado, is not above the same constraints and foibles that bedevil other studies. The RCT is not a knight in shining armor, but a gallant figure subject to the same trials and tribulations as its more scholarly counterparts.

RCTs A History

RCTs have their roots in clinical and epidemiological studies. This is perhaps its first impediment to their use in economics and the social sciences. Social issues are often more complex and have more than one causal pathway as opposed to drugs which usually have one casual pathway and a very easily verifiable target (a bacteria or a virus). The second impediment is that while both medicine and social sciences often use overlapping terms they often use quite different language when referring to RCTs. Thus what is known in medicine is not often known and and is almost never salient when considering an RCT in economics/ social sciences. We consider two issues :

• Average Treatment effect and why they are not the truth

• How to use an RCT’s results once we have them

Bias and Precision

Let us clarify two important terms: bias and precision. To a non-technical audience, the term "unbiased" often carries an unusually high status, perhaps because it is commonly associated with impartiality in political opinions. However, in statistics, being "unbiased" simply means that on average, the results are correct. It does not imply accuracy in every instance. Each individual RCT might produce highly erroneous results in either direction, but these errors tend to cancel out when averaged. Consequently, the fact that an RCT is unbiased provides limited value.

The second term, precision, refers to the degree of correctness on average. In the context of RCTs, precision indicates how close the results are to the true value on average. RCTs are notoriously imprecise, as illustrated by studies that have documented large errors in both directions, such as those involving hormone replacement therapy (HRT). This lack of precision is well-known, and economists often seek to enhance precision by incorporating "biased" and "subjective" expert opinions.

Measure Theory and the ATE

All misunderstandings about probability come from the confidence of individuals who have never had the wind knocked out of them by measure theory. And so in this section that is what we will do.

Fundamentally, the treatment effect, ’s equation is given by, Where the boolean variable is or accordingly as whether the th individual is in treatment or control. Ideally we would like to measure . That is, we would like to observe the same individual in treatment and control and measure the difference in outcome in the two cases. In absence of this we can only observe i.e. the difference in population mean between the treated and un-treated population. It is a remarkable theorem from statistical theory that says that, the difference in these means is an unbiased estimator of the treatment effect. This is remarkable because it requires very few if any assumptions. Recall, that unbiased-ness buys us relatively little for a study done once as it could be a completely random effect we observe in any one study. Below, is a measure theoretic proof of why this is the case, Where this last equation follows from the linear nature of the expectation operator. This is another vital weakness of the RCT, you can only ever get at the mean effect. You cannot get a meaningful measure of any other statistic. As an economist we are very often concerned with the median and below is a similar proof of why this is not the case, one can immediately see that this is a lot hairier to linearly separate than before, i.e. This is another critical weakness of an RCT, you can only tell what the treatment effect is in expectation. Though not useless, this is far from usual when a statistician would rather know the entire distribution of outcomes. This is generally the case with other forms of studies.

Randomization

Randomization is often looked at as this perfect tool that answers all questions related to variance between the treatment and control group but as we will see this is often not the case in practice. Recall,

Usually, the second term on the right is equated to . But there is considerable slight of hand involved here. While un-biasedness guarantees that the second term is in expectation. In any one trial we have no idea about the size of this term. This is referred to in the clinical trials literature as random confounding or realized (as in one realization of a trial) confounding.

Randomization and Balance

Exactly what randomization does is lost in popular parlance. There is often a misconception that randomization (in the sense of a laboratory clinical experiment) does as much as a perfect control. This is not the case, randomization is often always far worse than a good control. This fact is often lost in popular literature and can be captured by this quote in a World Bank manual attributed to Gerter et al 2016. We can be confident that our estimated impact contributes the true impact of the program, since we have eliminated all observed and the unobserved factors that might plausibly explain the difference in outcomes. This statement confuses the fact that the second term is zero in expectation over many hypothetical trials (which this study did not do) and with it being zero in any one trial. Popular economics literature is littered with such statements. It is this lack of nuance (and grace) that is the cause for RCTs being as widely misunderstood as they are.

Post Mortem

While, I have many misgivings about Bayesian epistemology in general, I think it is a very narrow way of viewing the world. It is also pseudo mathematical for a variety of reasons (see Pollock for a mathematical discussion of why it is not a tractable mathematical theory, but rather a subjective philosophy that appropriates real mathematics). For the reasons mentioned here, I think that RCTs do not often contribute enough signal to update probabilities about hypotheses in a meaningful way. In addition, it has become fashionable to quote a counter intuitive or counter-theoretical result from an RCT to suggest that theory needs to change, rather than requiring more studies or more information about causal pathways to improve the body of scientific evidence in either direction. Finally, the idea that RCTs require little to theory is patently false. Since a good RCT requires a good control which often requires a good theory which in turn is subject to all the shortcomings of the human experience such as political bias and subjectivity.

No causality in, no causality out.

Words to live by indeed.

References

Critique of Bayesian Epistemology (https://johnpollock.us/ftp/PAPERS/Problems%20for%20Bayesian%20Epistemology.pdf)
Understanding and misunderstanding randomized controlled trials (https://pubmed.ncbi.nlm.nih.gov/29331519/)

As the season draws to a close, no matter whether we win the league or not, I’m happy with where Arsenal are as a club. Although, I’d better get my feelings off my chest before Championship Sunday.

I cannot help but feel that our left backs cost us the league. Football seems symmetric to the untrained eye but really it’s never been more asymmetric, and no position exemplifies that like left back. And no team has suffered because of their left back as much as Arsenal have. Zinchenko is our best offensive option, but comes with very clear (though sometimes overblown) defensive limitations. Z just doesn’t have the instincts of a defensive left back, in fact Z doesn’t have many instincts of a defender at all. On some European nights he delighted us with an inch perfect left footed pass that scythed through a low block but on other nights he was taken to the cleaners by a virtually unknown European right back. In England, he was profligate with his passing and rendered unplayable versus the top right backs in England (unlike left back, attacking right backs seem to be sprouting all over England like an unwelcome disease). Our other options are all giants of men, who in their hearts know not whether they can string a pass together with the same careless abandon that Z can. Kiwior and Tomiyasu are both impressive defensive specimens but their quality on the ball shows that they’re really centre backs in hiding. As Gary Neville said (I think), if you’re taller than 6 feet you’d sooner play basketball than fullback. Which is good news since our secret weapon, all 5 feet 10 inches of Jurrien Timber will be available for selection next season. Though Timber will have to fight off doubts that he too is just a smaller centre back with incisive passing , such as he never showed in Holland.

Arsenal is not unique in their search for a quality left back. The famine of quality players in the position meant that only Destiny Udogie (of all people) was the only consistently featured “true” left back in the top 6 in England. That Gary Neville’s selection of him to his Premier League Team of the Season was met with raised eyebrows should tell you all need to know about the position. Manchester City have effectively given up on the position, playing centre backs (Akanji, Ake, Gvardiol) at the position, I don’t know that they’ve had a true left back on the roster in a while. Due apologies to Gvardiol, a goal in Madrid does not a full back make. Andy Robertson stands out as the best left back in England (for some years now) and a true representation of what the position can become when played with the fluency and elegance of someone who has both a left foot, football intelligence and about 3-4 sets of lungs.

Across Europe things don’t look any better. If you’d gone to get a cup of tea during the one forward run that Mendy made during their dalliance with City you’d struggle to name Madrid’s left back. You’d struggle to name his replacement too. A certain Welshman could walk into the squad right now if he played his boyhood position. I’ve always had a soft spot for Os Navegadores’ Nuno Mendes, he’s an incredible player, much like Robertson, but Sancho kept him up at night and ran rings around him during the day. Alphonso Davies stands out at the best player at the position in the world, how ironic then that it was his goal that spelt the beginning of the end of our European adventure. If a Scotsman and a Canadian are the best at left back in world football right now, surely our Brazilian cousins and Portuguese ancestors must be positively rotating in their graves.

All in all, I think left backs have fallen off since the glory days of the “wing back”, we still miss you Roberto Carlos. Arsenal have always fielded some greats at the position, from the flamboyant English left back of my childhood, Ashley Cole to the Frenchman Gael Clichy. Honorable mention to Patrice Evra and Leighton-bloody-Baines. Maybe the position has an evolution coming or maybe a mass extinction to be replaced by soulless centre backs who couldn’t make it. Or maybe we’ll figure out a way to play strikers at the position so that when we have the ball they can beat the trap and run into crosses. Oh wait…never mind.

Main Interview

1. Please tell us your name, your MA program and year of graduation, as well as your current role and title.
   Francisco Mendes, MACSS Economics, 2022. I am currently a Senior AI Engineer at Renesas Electronics America Inc.

2. Can you tell us about your journey after MAPSS-how you secured your current role?
   I think I had a pretty long journey to get here, but the short version is that I was recruited by on campus Consulting right out of our program. I spent about a year in consulting but having already had 5 years of Consulting experience I decided it was time for a change. Having worked primarily in recommender systems and signal processing in my entire consulting career, I decided to look for a career in that realm and here I am.

3. Did you always know that you wanted to seek placement in the academic or professional realm? Did this journey change at all for you?
   For me, I was really committed to go into the PhD program directly after MACSS. I think the opportunity was there to do so provided grades and research interests lined up. After attempting to transfer internally I was offered a pre-doc position to strengthen my resume for application to the Econ PhD program. To me a pre-doc program was not a feasible route to a PhD, I have several personal misgivings about the pre-doc system in Economics (it encourages primarily students from wealthy backgrounds to enter academia, it is not a competitive wage market equilibrium etc)

4. How do you feel that the MAPSS-Econ/MACSS-Econ programs prepared you for the role that you are in now?
   For me, my prior education in pure mathematics and statistics was probably most relevant to my current role. Econ programs in UChicago are a great way to learn some of those mathematical skills that are transferable to other domains. If you are certain you want to transition out of academia, I would highly recommend taking classes that reflect the domain you are interested in.

5. Is there anything that surprised you during your post-graduate pursuits.
   I think the one thing that surprised me (and continues to surprise me) is what a closed circle economics is as an academic field and how far it is from the economic and social realities of America. While you are in UChicago it may seem like Economics is all pervasive field that impacts the lives of many Americans, but reality is quite different, most of what we study here is very distant from the lives of everyday Americans. Even fairly well informed people will struggle to name leading figures in academic economics and most academic economics is irrelevant to policy making.

6. Can you tell us any important advice you have for graduating MAPSS-Econ an MACSS-Econ students?
   The first thing I would say is enjoy your academic life, single-mindedly pursuing a field of study and answering a research question that motivates you. If you have decided that you want to try moving into industry then prepare accordingly. Either way take charge of your career!

Other Random Thoughts

Expanding on some of the answers above ended up being worthy of a subsection, please find these below.

Pre-doc Programs

Economics is probably the most competitive grad school program in America right now. The number of applicants to the number of spots is quite high. The pre-doc system is a way to get a foot in the door, but it is not a guarantee of a spot in the PhD program. I think the pre-doc system is a cheap way to get labor for the university while parading itself as a way to get a better signal of ability. A pre-doc is no better signal of ability than grades and the Master’s research thesis. I believe if you are from a background that allows you to stay out of the job market for 2-3 years in your early twenties, then the pre-doc route does not cost you much. If you are older or have financial constraints, I would recommend against the pre-doc route. Pre-docs add, at best, a marginal signal to your application (outside of UChicago) because no one knows what you did. If you are not given a PhD at the place you did your PhD, other universities will look at you like a lemon (if you are not good enough for your own university, why would we want you?). In fact UChicago does this with pre-docs from Stanford and Harvard, they are not given PhDs at UChicago. It is very rare for a pre-doc to substantially increase their chances of a PhD at any place other than the place they did the pre-doc.

Transitioning to Industry

I think the most important thing to do is to take classes that reflect the domain you are interested in. If you are interested in AI, take AI classes. If you are interested in finance, take finance classes. Most of the economics program prepares you for one thing, a career in economics. Do not mistake the difficulty of the program for its usefulness and transferability. While the mathematical and statistical skills you learn, definitely make you better than a plug and play data scientist, they are not enough to get you a job in industry. You need to send a strong signal to the market that you are interested in a particular domain and have immediately applicable skills. The code you write in your classes is not enough to signal this. You need to have a portfolio of projects that you can show to potential employers. These days the coding requirements for most AI roles are quite high, you need to be able to code in Python and C++ at a minimum.

Economics as a Field

None of what follows is meant to be a coherent argument for or against the field of economics or the flavor of economics at UChicago. I think there are many benefits to both. What follows is simply a summary of my observations about how different conversations with fellow econs are versus everyday people (Americans for the most part). I had the habit of names dropping famous economists to make a point in everyday conversation, such as “as Milton Friedman said…” or “as Gary Becker said…” or “as John List said…”. I quickly realized that no one knows who these people are. I think the field of economics is quite closed off from the rest of the world. The field is not as influential as it thinks it is. That is for both better and for the worse, I think the field should be more relevant and more engaged with people outside the field itself. But in many ways it is for the better, I think that while you are in economics (especially at UChicago) you do not realize how ideological some of the classes are. For instance, price theory is really an ideological class that is not grounded in reality. Some of it might be true, but large portions are ideologies and assumptions about how markets work. When engaging with fairly well-informed individuals about economics, I realized that the goal of research at a deeply tactical level ends up being finding a niche and churning out papers in that niche. This in itself is not a bad thing, but if you zoom out (even a little bit) from that niche you realize that the research is not relevant in everyday conversation. I can think of two great examples of this. The first is any experimental behavioral economics research study, while many of the conclusions are interesting and unique, they remain just that interesting and unique. They are not generalizable to the population at large. And of course economists are the first to admit this, but this is a much bigger impediment to everyday conversation than they realize. I guess the main misgiving people have is that to the field of economics, every problem seems to be a nail (prices, wages, incentives) and the solution is always a hammer (money). Thus, every societal problem can be framed as one that can be solved using the right incentive structure. But this is a bigger ideological hurdle for most people than the field admits. The second would be the idea that markets work, this is purely a UChicago “fantasy”. Markets work in many cases and in many cases they do not. In either case, to the average American, the idea that markets work is not a given. While we spend a lot of time working out the math after establishing a set of assumptions, the assumptions we make are not as easy to establish to a jury of your peers, as it were. As an econ you are drawn to complex mathematical structures and their solutions (a matching is a lattice and a stable matching is a fixed point on the lattice, who wouldn’t be excited?!), the actual solution matters less than the assumptions and narrative around the problem.

Post Mortem

I enjoyed my time at UChicago and definitely enjoy the field of economics. I think the program offered me a lot of opportunities to grow and learn. I think the program is a great way to learn some of the mathematical skills that are transferable to other domains. But there are certainly some things you need to do (on your own) to get the most out of it professionally. Especially if your previous degrees are not relevant to industry.

Introduction: The concept of free will has long been a subject of profound philosophical inquiry, engendering discourse, and contemplation over the centuries. It occupies a pivotal position in our comprehension of human agency, moral responsibility, and the quest for a meaningful existence. This essay delves into the intricate tapestry of debate surrounding the concept of free will, examining the perspectives of determinism, the enigma encapsulated by the “hard problem” of consciousness, the role of randomness, and the computational attributes characterizing the human mind.

Determinism and Causality: Determinism, as a philosophical viewpoint, posits that every event, encompassing human actions, is intrinsically linked to antecedent events through the unerring chain of causality. From this perspective, the precise modeling and prognostication of the cerebral and neural processes would seemingly challenge the existence of free will. This notion finds reinforcement in the observation that the causality underlying many of our actions can be traced back to physical inputs, including cerebral states and external stimuli.

The “Easy Problem” and the “Hard Problem” of Consciousness: To elucidate this perspective, proponents of determinism segregate the dilemma into the “easy problem” and the “hard problem” of consciousness. The “easy problem” is concerned with deciphering the physical mechanisms underpinning our conscious experiences, suggesting that a successful resolution might pave the way for deterministic explanations of human conduct. However, the “hard problem” of consciousness, as delineated by the philosopher David Chalmers, introduces a more profound quandary. This problem revolves around the query as to why and how the physical processes within the human brain give rise to subjective experiences, thereby introducing an element of mystery and insusceptibility to reduction.

Challenges to Determinism: The “hard problem” of consciousness introduces a formidable challenge to the principles of strict determinism, requiring that consciousness and subjective experiences transcend facile reduction to mere physical processes. Furthermore, the arguments posited against determinism offer significant insights into the ongoing debate.

The following essay attempts to summarize prevailing notions of free will and reasons to believe in its absence. This is not meant to be a summary of current philosophical stances on this debate, nor is it meant to be an authority on the subject itself but rather it is meant to over emphasize arguments for and against the issue that stand out in my mind as relevant based on our discussion, in particular, the tone of this essay is meant to be scientific and not philosophical, with issues surrounding the metaphysics of the debate largely left to the readers own research.

1. Weight of Intuition: It is posited that, in the absence of empirical evidence establishing a concrete link between physical processes and human conduct, a measure of importance must be ascribed to the intuitive sentiment of free will that is experienced by all human beings. This underscores the concept that, pending further scientific exploration, one ought to remain ambivalent before conclusive determinations concerning the presence or absence of free will are made. Intuition, while not constituting an unequivocal argument, functions as a reminder of the unfolding progress in our comprehension of complex phenomena. Human intuition does not (in and of itself) run counter to scientific reason, it is the guiding principle against which scientific accuracy is verified. If human intuition does not align with a scientific theory, great care is taken to experimentally verify the theory. Thus, human intuition is a motivating factor, not a limiting element for scientific discovery. Human experience not aligning with scientific theory drives scientific enquiry, decrying the human experience as an illusion to fit scientific theory (which itself might be at a nascent stage) can be counterproductive to furthering scientific enquiry.

2. Complexity of Modeling: This discourse delves into the intricacies involved in modeling systems with a plurality of variables, emphasizing the critical distinction between modeling a solitary variable leading to a definitive outcome and modeling a multitude of variables collectively influencing an outcome. If human behavior can be reduced to modelling inputs and outputs, then it is fair to say that a comprehensive model of human behavior would end the notion of free will. The evidence does exist for this to be true, multiple studies have shown links between physical processes in the human body and behavioral outcomes. Most notably, the case of a man developing a taste for child pornography due to a tumor in the frontal lobe. Arguments against generalizing this can certainly be made, but here we seek to show that even if such a model and such input output mappings can be established, the complexity of such mappings is a non-trivial issue, since one such complex mapping is “free will”. Thus, simply saying there “could be” a complex mapping is not enough, some bounds must be imposed on the mapping for it to be considered a viable and testable scientific theory (indeed if one does not place such a bound of complexity, one runs the risk of indulging in scientism). It is astutely observed that the simplicity of a model is an essential criterion for it to be recognized as a viable theory of the functioning of the world. In practice, models featuring a multitude of variables may render assessments unfeasible, thereby rendering it arduous to confidently pronounce the presence or absence of free will within such systems. For instance, let us say we have a model for whether someone buys apples based on their blood sugar, Apples = 3 * blood sugar + epsilon. Then for sufficiently large numbers of people, we can observe the data (Apples, blood sugar), posit a model, and verify its accuracy. Suppose that this fails, suppose we are unable to create a model for apple purchases, Apple = f(blood sugar) + epsilon. We then can introduce more and more variables, Apples = f(blood sugar, weight, color of the flag of Mexico, …., day of the week), at some point along this journey the problem does not become that of whether one has the right model or the variables, it becomes that of whether the process being modelled can be modelled at all, as the amount of data required to verify this model increases exponentially and the theory might never be tested. Thus, a highly complex model would be indistinguishable from “free will” and indeed the latter might be a simpler explanation than hypothesizing a complex model without experimentally verifying it.

3. Role of Randomness in Subjective Experience: A contrary perspective asserts that subjective experiences might be due to random occurrences amidst the sequence of physical processes, yet this does not intrinsically proffer evidence for the existence of free will. This viewpoint underscores the notion that the mere recognition of randomness does not necessarily imply the existence of free will. In this context, randomness might be construed as a contributing factor in the intricacies of human conduct, sans signifying genuine free will. Here again, we can take recourse to models of human behavior, if Apple = epsilon i.e. the model of apple buying seems to be truly random, then it is also indistinguishable from free will. Thus, randomness of human behavior could certainly be due to a lack of free will or an important empirical consequence of it. It certainly raises the bar for proof of the absence of free will. Any theory of Free Will should necessarily disentangle randomness from Free Will.

4. Computational Nature of the Human Mind: The perspective articulated herein underscores that the essence of free will resides as the most distinguishing feature of the human mind. The contention posits that contemporary computational devices remain bereft of the emergent property of free will, thus demarcating the human mind as distinctive. The counterpoint presented by an opposing viewpoint underscores the core notion that, notwithstanding the intricacies and emergent properties inherent to the human brain, it ultimately finds its basis in physical processes. This reductionist standpoint emphasizes the primacy of physical processes as the fundamental foundation upon which human cognition and conduct are predicated.

6. Distinguishing Worlds with and without Free Will: A thought-provoking analogy is proffered to discriminate between a world with free will and a world bereft of it. In a world governed by free will, human actions assume an unpredictable character, rendering the task of forecasting the movements of individual humans a challenging pursuit for an external observer. This analogy accentuates the complex and unforeseeable nature of human conduct in a world animated by free will and underscores the formidable challenge of disentangling unpredictability from randomness.

7. Free Will and Science: Neither viewpoint, whether it’s the belief in the compatibility of free will with the laws of physics or the belief in their incompatibility, is inherently more “scientific” than the other. The compatibility of free will with the laws of physics is ultimately a philosophical and metaphysical question, and it falls outside the realm of empirical scientific inquiry in the traditional sense.

The scientific method is primarily concerned with empirical observation, experimentation, and the formulation of testable hypotheses. It is well-suited for investigating and understanding the natural world and the physical laws that govern it. However, the question of free will, particularly when it pertains to its compatibility with the laws of physics, involves conceptual and philosophical issues that go beyond empirical observation.

Scientists may conduct experiments to study decision-making processes, brain activity, and related phenomena, but the interpretation of these findings in the context of free will remains a matter of philosophical interpretation. Some scientists may hold philosophical positions on the matter, but those positions are not inherently more scientific simply because of their scientific background.

In summary, the question of free will’s compatibility with the laws of physics is a philosophical and metaphysical inquiry, and it does not fall strictly within the domain of empirical science. Scientific methods can inform this debate by providing insights into decision-making processes and neural activity, but the question itself is inherently philosophical in nature.

Conclusion: In summary, the question of whether free will is a tangible phenomenon remains unresolved. The deterministic perspective, grounded in the principle of causality, offers a compelling viewpoint, suggesting that an in-depth understanding of physical processes may limit human agency. However, the scientific landscape on free will is dynamic and inconclusive.

Recent neuroscience advancements have shed light on the neural processes underpinning decision-making, indicating that actions can be predicted based on neural activity before conscious awareness. This sparks debate on the extent of conscious control over choices, yet it doesn’t provide a definitive answer.

The debate encompasses philosophical, ethical, and psychological aspects, with some arguing that neural underpinnings don’t negate free will if choices align with desires and values.

It is essential to note that while recent advancements in neuroscience may suggest neural pathways for simpler decisions, this doesn’t necessarily apply to more complex decision-making processes. For example, while hunger might correlate with increased spending in a shopping mall, it’s not reasonable to assert that someone’s overall spending patterns can be directly linked to the food they eat. This highlights the intricate nature of human behavior and decision-making, further emphasizing the complexity of the free will debate.

Free will is a multifaceted subject, challenging our understanding of the human experience. It requires sustained scientific and philosophical inquiry to unravel its intricacies. For the most current insights, consult recent scientific literature, recognizing the evolving and interpretive nature of this topic.