Gaming the Harvard UC Grant System

And they say arbitrage doesn’t exist in the real world

When your club event gets funding for all 25, 000 participants! (source)

Who needs Chaucer? Harvard’s student organization events are so good that 37.3 out of every 10 students attend them. Overestimating attendance is advantageous for clubs, which receive UC funding for food, transportation, and instruction on a per-student basis. A follow-up report in October indicates that the UC Finance Committee — FiCom for short — is still looking for ways to confirm club estimates of attendance and spending.

In the meantime, anarchy? It’s true that FiCom recently instituted a “New Online-system for Vetting Applications” also known as nova, which recently suffered a hacking attempt. Could it be from some foreign entities wanting some funding for their own highly-attended events? (FiCom says no: it was just a generic bot attacking any website that it saw.)

Either way, for the less hacking-inclined, this article discusses UC grant data from the 2015–2016 school year and how clubs can use certain trends to their advantage:

  • Part 1: “The early club gets the grub” discusses temporal trends
  • Part 2: “Presents of past and future” compares upfront vs. retroactive grants
  • Part 3: “The best spenders” features the clubs who received the most funding in 2015–2016.

Keep note: This is a report on the 2015–2016 school year, and FiCom funding policies and practices have changed significantly since the initial analysis was run. Stay tuned for a 2016–2017 report later this school year :)


Part 1: The early club gets the grub

Some quick stats about the 2015–2016 school year:

  • There were 882 requests totaling $551,430.40.
  • 48% of the requested amount, or $327,366.22 was granted by the UC.
  • Individual requests ranged from $0 to Veritas Financial Group’s whopping $7434.50. (Can’t they finance themselves, or something?)
  • Almost all of the requests (845/882) were under $1000, and 30% of requests (256/882) were between $50-$100.

Examining the ratio of allocated : requested — that is, how much the club received vs. how much they asked for — brings us to an interesting comparison of ratio against grant number, shown below. Grant 1 is the first of the school year, and grants with higher numbers are given later in the year.

Comparison of grant allocation ratio and grant number

A number of properties stand out:

  • Before grant ~350, the UC almost never gave clubs $0, whereas after grant ~350, the UC gave $0 quite frequently.
  • Three “blocks” appear after grant ~500, in which the UC granted approximately the same percent of the requested amount to numerous clubs.
  • After grant ~525, the UC almost never granted the full amount of the request.

FiCom only grants $0 to clubs that make a mistake on their application or fail to show up for their grant interview, so they attribute the increase in $0 grants to clubs that run out of time to interview at the end of the semester. Dialogue with FiCom also suggests that they scale grants down proportionately when the weekly budget is exceeded.

It is then advantageous to:

  1. Plan pricier events for the start of the school year to allow for more time to interview
  2. Interview for UC grants when other high-cost events are not competing for the same funding

Part 2: Presents of past and future

Among the 882 requests, there were four types of grant: Bridging and Belonging, Large Venue Grants, Club Sports, Retroactive, and Upfront. (Note: as of Fall 2016, the Large Venue Grant has been succeeded by the Grant for an Open Harvard College.) A retroactive grant is given after an event has been paid for by a club, and an upfront grant is given before the event has been paid for.

Total discrepancies are a result of missing data. We might expect retroactive grants to have a higher ratio for a number of reasons:

  • Clubs asking for money after an event take on more risk, so they should tend to spend more conservatively. Because they can do so according to the UC policy guide, retroactive requests should have a higher allocation : requested ratio.
  • Retroactive requests are limited by how much a club spends, whereas upfront requests are limited by the club’s vision. If a club has an idea for an event that costs $1000, and a super-version of that event that costs $1500, then it’s beneficial for them to ask for $1500 ahead of time instead of spending $1000 and realizing they could have gotten $1500. (But before you get too excited, clubs must return money that they don’t spend.)
  • Asking for more money upfront has positive expected value, so clubs should be willing to ask for more money through upfront grants, and upfront requests should have a lower allocation : requested ratio.

However, there is no statistically significant difference between the means of Upfront vs. Retroactive ratios, so retroactive and upfront grant applications seem to be equally successful.

It follows that clubs should use upfront grant requests as often as possible and overestimate expenses, minimizing the risk of spending more than they are allocated.

Meanwhile, the UC should continue to keep track of club expenses through photographed receipts and internal verification to ensure that money for upfront grants is used properly!


Part 3: The best spenders (aka, the UC’s finest sugar babies)

The same five student organizations consistently requested and received the most amounts of money. The Harvard-Radcliffe Dramatic Company takes the leaderboard with 28 requests totaling $24,175 — perhaps this will inspire other clubs to apply for grants more aggressively.

Behind these top 5 in allocated amount are IvyQ, Harvard-Radcliffe Orchestra, Bach Society Orchestra, Harvard Club Baseball, and Harvard Cheerleading. Out of the 20 clubs receiving $3,000 or more last year, 9 of them were athletic organizations, 4 were arts-related, and 4 were cultural organizations.

According to the policy guide, “Approved club sports teams will receive a maximum of $1,000 — $2,000 in UC funding per semester.” Many sports teams have come close to their maximum, whereas others do not take advantage of the maximums described in the policy guide. Read the guide — free money awaits!


Part 4: “What do you mean?” -Justin Bieber

To recap, clubs should:

  • Plan costly events for the beginning of the year, and avoid planning big events at the same time as other clubs
  • Take advantage of upfront grants, which are equally as successful as retroactive grants
  • Read the policy guide to receive the maximum amount of money for food, transportation, and more
  • Stay tuned for the 2016–2017 school year report!

In the spirit of open data, feel free to take a look at the FiCom records or the data and code used in this project if either piques your interest.

Meanwhile, thanks for reading and may all of your club events be impossibly popular!


This article is one in a series of data analysis projects that the Harvard Open Data Project conducted in the Spring 2017 semester investigating Harvard’s open data sets. It was originally a report written for the UC Finance Committee.

The Harvard Open Data Project is dedicated to opening and analyzing Harvard’s data to empower members of our community to improve campus life.