I. Introduction: Exploring the Unexamined

In a lot of ways, the work for this project is inspired by, and is meant to build upon, the research and results achieved by The Equality of Opportunity Project (EOOP). The EOOP includes data on the economic and social mobility of students from hundreds of schools across the nation. Our group has gathered data on the 20 schools with the highest mobility rate. We hope to compare these top 20 schools and see what characteristics they have in common. Perhaps if we study them closely we can figure out what it is that makes these schools so successful at moving students who come from the bottom 20% of the nation up to a higher earning potential.

As students at Emory University, our immediate interest came at the college level, where we were able to see which schools performed well and which didn’t quite stand up to par. Amongst other elite colleges, Emory enrolled the second-highest amount of low- and middle-income students (narrowly behind list leader UCLA), but beyond that, performed relatively mediocre, especially considering the depth of its endowment and the ‘prestige’ some associate with our school. In many ways, the results of the EOOP’s research bring to question many of the standards we hold of our schools and what it is that constitutes a school as ‘good’. Should mobility rate, the percent of a school’s students who have parents in the bottom 20 percent of the income distribution that succeed in reaching the top 20 percent of the income distribution, be a major element of the college evaluation process? Are the Ivy League schools, and even schools like Emory, really as good as we claim them to be? Or are we simply evaluating them on certain criteria that allow them to stay within America’s upper echelon of perceived quality? The EOOP left us with these questions, and we wanted answers. Below, you can see the top 20 schools in order of descending mobility rate.

Immediately, one can see that the schools performing best in the mobility rate column (meaning they achieved the highest percentages) are not schools that anyone would have been able to guess, or even ones that the majority of people would know, for that matter. In terms of tier, 10 of the top 20 are selective public. Only one of the 20 falls within the category of ‘highly selective,’ and none within the Ivy or Elite categories. Clearly, the schools that typically receive praise and recognition for being the greatest academic institutions are not evaluated heavily on their mobility rate. Whether they should or not is a discussion for another day, but regardless of mobility rate’s worth in comparison to other evaluative techniques, it is a measure worth investigating. Why do some schools perform better than others in terms of mobility rate? Are there any clear factors that drive success in mobility rate? What makes the top 20 schools in mobility rate stand out from the pack?

With these questions in mind, our group began our investigation. In this section are analyses on college characteristics, including location, race, major distribution, and cost. While these analyses are only the beginning of the story, they will hopefully offer a basis by which we can begin to understand what, if anything, sets the top 20 apart.

State Distribution

Here, we'll be taking a look to see if there is some correlation between location of these colleges, accesses, financial availability and fund provisions for these top 20 colleges. We’ll be considering various factors to find similarities and dissimilarities between the states. Maybe region contributes to where the top mobility colleges are located where they are and why they are so successful? Maybe there are external variables missing or contributing? Although finding a definite correlation between college location and upward mobility is unlikely, we still find it interesting to compare the top 20 colleges in this way.

The data presented in these graphs were obtained from Equality of Opportunity College Level Characteristics. Graph 1 compares the location of the top 20 colleges and graph 2 compares the locations of all 190 colleges from the data. The top 20 colleges are highlighted in a separate graph for they had the highest mobility rates out of 190 colleges in the data. Looking at graph 1, 75% ofthe top 20 colleges are in the northeast, 15% are in the south and 10% are in the West. These top 20 ranking colleges in the data set provide the highest upward mobility of students enrolled. These colleges are less selective than Emory for instance but their mobility rate exceeds Emory along with other top ranking colleges like Harvard, Stanford, Duke, etc. Here, we'll be taking a look to see if there is some correlation between location of these colleges, accesses, financial availability and fund provisions for these top 20 colleges. We'll be considering these factors to find similarities or dissimilarities between the opportunities provided regionally and how that contributes to where the top mobility colleges are located where they are and why they are so successful.

Table 6 provided by the National Student Clearing House Research Center had a data table listing the estimated enrollment of all 50 states. Again looking at this data, California, New York, and Texas have the highest listed enrollment numbers for the 2014, 2013, and 2012 years. However, it is possible that state population is a lurking variable in the data. According to the 2010 census Texas, California, and New York are highest populated states in the US.

May Wong wrote an article for Stanford News titled Stanford research analyzes colleges as engines of upward mobility. The article mentions that public colleges and universities “dominate… highest mobility rates, and all are located within California, New York and Texas”. Data collection for that research study, was gained from studying the “anonymized tax records” of 2187 American colleges from years 1999 – 2013. The data used to generate graph 1 was obtained from Equality of Opportunity Project. It’s not surprising to see the top states having the highest mobility (out of the top 20) being the exact same states mentioned in this article.

The article then goes on to describe how these colleges admitted have large numbers of students who are low -income but they’ve yet to see how well and what is the correlation between mid tier public schools and high mobility rates. The article also mentioned a paper from Equality of Opportunity Project which talks about how kids from richest/top 1% class are close to 80% likely to attend Ivy League colleges compared to the nations bottom 5th percent. This article does not articulate the connection between state and mobility. It’s difficult to find the connection because a multitude of external factors are at play. The community, surroundings, funding, and public school education of states are not entirely equal so it’s difficult to control for that and pinpoint exactly why there could be an association or trend.

It's also interesting to look at the top states which have the highest economic mobilitiy. According to the Pew Economic mobility project, only New York (out of the top20 college states) makes the list. Texas scores worse on the mobility scale and California does not have a statistical difference from the national average. According to this figure, a couple of states from the northeast provided the "better mobility" than the national average indicating there is something about the North East. However, when you refer back to graph 1, NY, TX, and CA are the top 3 but NY by far has the most top 20 colleges located in it.

Major Share Distribution

We wanted to study the discrepancies in major distribution between the top 20 schools and those outside the top 20 because some majors are statistically proven to lead to more financially rewarding career paths. We looked to see if it was just a simple case of certain schools devoting more of its resources on academic fields with the highest financial potential.

As you can see in the bar graph above, compared to the data set of all colleges, the top 20 schools’ aggregate major distribution is dominated by STEM subjects as well as majors related to public/social work. There is a +8% percent change in the number of STEM students compared to the average major distribution of the schools outside of the top 20, as well as a +5% change in public/social work majors. Moreover, social sciences and trade and personal services are also attractive options for students who attend the top 20 schools. Perhaps not too surprisingly, there are fewer students in these schools who pursue the arts or humanities subjects, but in the same vein, I would not have expected the social sciences to be such a popular field. Also, there are fewer Business majors in these schools, which was also a bit of a surprise to me. Overall, I think the overall conclusion that can be drawn from this data is that students in the top 20 are more drawn to hard skills and the technical fields. Whether it is because these schools offer attractive programs in these departments and do a better job recruiting students who are interested in these subjects cannot be determined in this dataset.

Race Share

Dylan Matthews article in Vox talks a lot about how important mobility rate is for American students seeking the American Dream. In his article he states how important it is for elite schools like Harvard or Stanford to not take so much pride in elevating small groups of financially underprivileged students when other schools have a better record of doing so in sheer numbers.Here we task ourselves to look not just at why this lack of access creates a detrimental system for youth in this country, but also why the schools that have such a high mobility rate are so successful.

How does race play a role in this matter? In order to explain that, we can look at diversity. If we take the top 20 schools in mobility rate, one item may catch your eye:

See it? No? Let me highlight it.

14 out of the 20 schools with the highest mobility rate are in the State of New York. In fact, if you look closely, 8 of those 14 are actually apart of the same school system, The City College of New York. Does this mean that "New Yorkers" are just better at creating programs to move students into the top 1%? Perhaps, but lets take a look at the numbers.

So why are we even referencing state when the title of this section is clearly "Race"? Here's why: take a look at the following comparison between the percent change in different ethnicities at these schools.

The chart above shows how the top 20 schools fair in each category of ethnicity. Obviously these schools have a significantly higher amount of minority students applying, getting an acceptance letter, and attending the school. Don't believe me? I'll let the numbers speak for themselves. The following shows CUNY's Student body make up by high school:

Crunching the numbers shows how 92.2% of the student body is made up of students from NYC highs schools and 94.8% of them are from NYS high schools. Because the surrounding area is so diverse as seen in Figure 4, it is easy to see why the population of the CUNY school system is so diverse when this is where the schools get their students from. This diverse community inherently raises the amount of minority students in the these institutions. The following interactive shows how some districts in NYC have a low median income.

The sheer number of students who go to these schools must be statistically represented in the schools above. When these institutions are successful at moving students up into a higher financial tier, it is clearly better reflected. A higher number of underprivileged students in the applicant pool, leads to a more diverse(both racial and economic) student body which leads to higher mobility for the schools we see in the top 20. The above remains true for the student body of the other schools in NY, TX, and CA.

Scorecard Net Price

The cost of a college education may be the most intuitive indicator of mobility rate success. Naturally, students from low-income families will have a better opportunity of success if they are able to afford college. If you have the adequate funds, one’s odds at attending college go up, and consequently one’s likelihood of graduating follows suit. Unfortunately, determining which schools are the most affordable is not as simple as it sounds. Each school has its own net cost, but between scholarships and financial aid, cost can vary greatly not just from school to school, but from student to student. This case-by-case cost structure is in large part why so many have tried to determine which colleges are the most affordable for low-income students and what it is exactly that makes some colleges more costly than others. The EOOP did some research of their own, providing what they call the ‘scorecard net price,’ the net cost of attendance for the bottom 20 percent income quintile. This is essentially a way of evaluating how costly it is for low income students to attend a given school, helping simplify the problem of determining cost on a per-student basis (to an extent). We looked at the data for 2013 and attempted to compare the scorecard net price for the top 20 schools with the data set of every school in the EOOP’s study, as you can see below.

Looking at the two histograms, the first thing to note is an obvious problem: the axes of each histogram vary, as do the bin sizes for each histogram. Thus, it is difficult to compare the data between the two charts, particularly if you tried to superimpose one histogram on top of the other. However, the histograms still provide us with a reasonable basis of understanding each data set.

What stood out the most to us was the second bin in the Top 20 chart, resting between $3,000 and $6,000, which towers above the other bins for that histogram. 10 of the top 20 schools rest in the bin, indicating a potential correlation between mobility rate success and scorecard net price within that range. Moving over to the histogram for all colleges, the largest bin lies in the $6,000 to $7,000 range, with a lot of the data tailing off to the far right end (more costly) of the histogram.

Building from this information, we looked further into the data to see whether the apparent tendency towards lower cost for schools successful in mobility rate shown in the histograms was a legitimate statistic. In the 3-6k bin in the top 20 histogram, the costliest school was $5,148, meaning 60% of the schools in the top 20 list had scorecard net cost below $5,148. Translating this value over to the entire data set, only 12.6% of schools had a scorecard net cost less than or equal to $5,148.

With this in mind, it is clear that those schools performing well in mobility rate tend to provide a significantly lower cost to low income students than the majority of their peers in the complete data set. Thus, for those schools hoping to improve their mobility scores, the data indicates that aiming for the lower end of the scorecard net price may be a factor worth pursuing. There are a number of schools with higher scorecard net price, most notably MCPHS University which has a scorecard net price of $28,944, so we know that there must be other factors at play besides cost if schools nearer the upper end of scorecard net price can still succeed in terms of their mobility rate. Caveats aside, the large percentage of schools in the top 20 that possess particularly low scorecard net prices provides strong evidence that there is a correlation between scorecard net price and mobility rate.

VI. Conclusion: Some Answers, More Questions

In terms of geography, ¾ of the top 20 colleges with the highest rate of mobility are concentrated in the northeast. In fact, 14 out of 20 schools on the list are located in the State of New York, and the CUNY system (The City University of New York) makes up 8 of the 14. However, because of many other variables that are either unquantifiable or hard to find data for, such as funding and public school education of states, it is difficult to spell out the connection between state and mobility.

As for the major distribution, we did find that as a very general finding, the top 20 schools attracted more students in STEM subjects and also public/social work.

With regard to whether race plays a role in determining one’s mobility rate, we found that these schools tend to recruit a higher amount of minority students than schools outside the top 20, with Hispanics making up almost 25% of the individual student body of these schools.

Lastly, as far as the net cost of attendance, we can see that schools that have a high mobility rate also seem to carry lower cost to low income students. A suggestion for the costlier schools would perhaps be to try lowering their scorecard net price.