This is the next in a series of posts by recipients of the Career Services Summer Funding grant. We’ve asked funding recipients to reflect on their summer experiences and talk about the industries in which they’ve been spending the summer. You can read the entire series here.
This blog is by Iris Zhang, CAS ’16.
I spent this summer as an economics and statistics intern with the civil rights bureau of the New York State Attorney General’s Office. As someone who grew up in Hong Kong, I felt incredible nostalgia to be back in a big city, although my personal biases dictate that New York City doesn’t even compare to one-tenth of how amazing Hong Kong is. The civil rights bureau, though, that’s something else. I am a rising junior, and ever since freshman year when I took the Benjamin Frnaklin Seminar, “Race, Crime and Punishment”, with Professor Marie Gottschalk, I became incredibly driven to civil rights issues, in addition to my passion for women’s rights. After spending my freshman summer interning with the Institute for Women’s Policy Research in D.C., I became highly aware of the important role numbers play in impacting policy. Discouraged by the political climate in which not much gets done regardless, I was interested in the use of numbers and statistics to make forceful legal arguments – the legal arena, it seems, is a much more effective (though not always more efficient) way of ensuring our rights and liberties through enforcement.
The civil rights bureau was the perfect place for me to test this instinct – I met some of the most passionate and talented attorneys who use their gift for a truly important cause. My role as an economics/statistics intern was to analyze large sets of data to extract meaningful statistics that indicated evidence of discrimination in aid of the bureau’s various legal investigations. I started the summer hoping to find out more how lawyers build numbers and statistics into their arguments, and now that I have tearfully bid farewell to the summer, I have come to a few conclusions about statistics and the law.
1. The application of statistics can be hard!
This is probably because I am neither a lawyer nor a statistician. All of the sets of data that I worked with were completely new to me (and in some cases, new to the attorneys too). The problem with this is I didn’t quite know how and where to start. You have the data, but this might be a new type of litigation or a new type of data, and you go into it not even knowing if there is anything meaningful contained within the numbers. For example, there was a discrimination case we were working on, and we had this data set with well over a million entries in the excel spreadsheet. The investigation had already cycled through 4 or 5 interns. I had never seen that data before or even heard about this type of legal issue. It was a lot of finding my way in the dark, and that in itself pushed me to want to learn more about statistics, because part of the learning process is developing the mindset to see a big body of data and think – okay, what can I find out? The lawyer’s mindset is surprisingly well-adapted to that. They would ask, “do we have information on how many people who were hired are minority applicants?” “Okay, so do we have information on how many people who applied were minority applicants?” “Do we know the demographics of the market pertaining to this job?” These are easy enough questions to ask if you are a lawyer because they are grounded in legal theory, and it was so interesting to churn out numbers for the attorneys to analyze. I was learning – okay, if you want to prove employment discrimination, the law says this – at the same time as I was learning whether the numbers complement that legal theory. But it was hard. Government bureaus don’t have that much money, so they find some 20-year-old hack (me) to try to make sense of the data, and I was really grateful for the experience.
2. Numbers lie
You can’t always trust numbers – they can be easily manipulated. In a discrimination investigation we conducted, we wanted to compare our target to its peers to prove that their discriminatory practices were yielding disparate impact, and statistically significant differences are a good way of supplementing observations. The numbers can lie if you alter what you define as “peer”. If you compare the target to their stronger, bigger competitors, they most certainly will fall short. The target wants to be compared against their weaker, smaller competitors to come out on top. So the question is – which peers do we compare them to? In this case, one really has to peruse the data carefully and base that determination on strong theory. But perhaps the better way to phrase this conclusion is to borrow the catchphrase from the NRA – numbers don’t lie, people do.
3. You can’t rely on numbers too much
Bouncing off the previous thought, if numbers lie, then you really can’t rely on numbers. On top of that, there can also be human error. For example, we had a data set that had total occurrences of a particular incident in 3 years, and also population information for 3 years. Do get the rate per unit of population, you can either use the total occurrences in 3 years divided by the total population in 3 years; or the average occurrences divided by the average population. In my initial analysis, I used the total occurrences divided by the average population, so the rate became way, way inflated. I was so embarrassed when I double-checked the data and noticed my mistake because it seemed so basic, but luckily, despite my frantic explanation, the attorneys didn’t seem to have an idea what I was talking about. This was a silly error, but on a more serious note, the experience solidified what a mentor had once cautioned me about using numbers. She said that numbers can certainly make a point, but numbers can be a convenient way for people to stave off having to make more important moral calculations. Even with no mistakes, on paper, the numbers might make it seem justified to take an action when it might not be the most morally sound idea.
I was extremely grateful to have had this learning experience, and it only served to solidify my interest in the application of statistics in law and social sciences. Without funding from Career Services, however, I would never have been able to afford the exorbitance of New York City.