This summer we worked with Professor April Beisaw on the community heritage and archaeology project that researched the histories of towns that were impacted by the New York City water system. Additionally, we responded to community requests for archaeological expertise and helped organizations such as Marist College and the Ashokan Center with surveys and excavations.

The majority of the project was spent doing a combination of archaeological surveys with community members and lab work to decenter New York City as the central part behind the history of the NYC water system.

The NYC water system is a complex system of watersheds, reservoirs, and aqueducts that bring water from rural populations in the Catskills/Delaware Watershed and Croton Watershed to city populations through the use of gravity alone. Although some claim this is a marvel of modern engineering, what is not told is the displacement of communities in order to create this system. However, evidence of this destruction can be seen by the building foundations that are scattered across the lands surrounding the reservoirs.

Through community hikes of the Neversink, Rondout, and Ashokan reservoirs, we attempted to use ruins as stepping stones that could spur stories of the region from locals. The most important aspect of this project came in the lab as we combined GPS data, historical research, and oral histories to create interactive websites, StoryMaps, that shared our research back with the community. Our experience with this project emphasized the importance of including the public in academic research. Not only would we not have known much of the information we learned without them, but by documenting their stories we were able to help center their local knowledge and reveal the history of struggle and perseverance that happened as a result of the displacement of their communities.

I worked with Professor Bart Thurber this summer to prepare for his spring seminar. The seminar will explore how museums have come to be as they are conceptually and materially—how renovation, reinvention, and, in some cases, demolition have played a role in their evolutions.

Our first and primary task was to get to know the eleven museums that would be used as case studies throughout the course. Ranging in geography, history, architecture, and design strategies, our list included the State Hermitage Museum (St. Petersburg, est. 1764), the Guggenheim (New York City, est. 1937), and the Louvre Abu Dhabi (est. 2017). I located as many digital resources as possible for each institution and immersed myself in them; I navigated museum websites, examined original floor plans, viewed 360 photographs, and read widely to get a detailed sense of each space in context.

This immersion segued into the second objective of our project: to develop materials for the course and flesh out its structure. From my digital explorations, I wrote introductions and curated “resource libraries” for each museum. The latter included webpages, journal articles, photographs, virtual tours, maps, archival documentation, and journalistic accounts. Some of these references will serve as core readings, while others will be tools for independent study.

While producing these materials, Professor Thurber and I worked towards distilling a list of shared features of museum institutions to be used as a framework for comparison throughout the class sessions. We also compiled general issues concerning museums and brainstormed ways the Loeb could find its place in the course as a “laboratory” for thinking.

I was glad to have spent time gathering resources, refining ideas, and architecting a learning journey. In many ways, those are the very processes out of which museums come to be.

This summer, I assisted John Murphy with preparing materials for a fall intensive class and curating the fall spotlight exhibition on the Women’s Studio Workshop. While both projects divided my time, many of the materials overlapped, allowing for cross-thought on how the content could engage each other within a class framework.

The Women's Studio Workshop (WSW) is a studio and arts space in Rosendale, NY. Founded in 1974 as a women-centered art space, and now, under new leadership, WSW has begun to tease out what exactly “woman-centered” means in a contemporary context. In approaching WSW as the subject of the spotlight exhibition, we were interested in responding to this question alongside telling a brief history of the Workshop. In using the spotlight space, we were also engaging Vassar's connection to WSW as a repository for all their artist's books, as well as engaging WSW's role in the broader Hudson Valley arts community.

In addition to this project, I also planned one class session for "Paper Protests," the fall intensive. After a visit to Interference Archives in Brooklyn, I became interested in the spatial theater that posters create when in a public space, and how that theater mitigates the original intention of the posters. As a community-centered organization, WSW has collected posters created for various programming and events since its inception. WSW became an outlet to explore spatial theater on a micro scale, while I found readings to guide a discussion of spatial theater on a macro scale.

Between all the research, I visited many locations across the Hudson Valley, including the WSW archives, Art Omi, and Bard CCS. I'm very thankful to John Murphy and the rest of the Loeb staff. I look forward to further exploring the curatorial process in future endeavors.

This summer I worked with Professor Argudo and four other interns to continue research on the wage and wealth gaps between people born in the United States (henceforth called natives), authorized immigrants, and unauthorized immigrants. Previous research on this topic has documented wage disparities between all three nativity groups and wealth disparities between natives and immigrants. Our research is novel in its distinction between types of immigrants when surveying wealth gaps, as well as the inclusion of asset holdings.

Our work built on Professor Argudo’s 2021 Ford Scholar project. We continued to use Survey of Income and Program Participation (SIPP) data, which changed considerably during the time period we examined. Since we did not finish matching the variables we needed last summer, the team and I began homogenizing the remaining relevant variables such as race, occupation, and industry from 1996 to 2020. This left us with a nationwide dataset of individuals that could be tracked over multiple months and years.

The initial stages of our analysis focused on using the homogenized data to replicate figures from previous papers that compared immigrants and natives to confirm our data was valid. Once confirmed, we distinguished between authorized and unauthorized immigrants using George J. Borjas's methodology, and split the data into workers of different education levels.

The final days of the project were spent starting a paper to summarize our findings and introduce our research, along with creating graphs on wealth holdings by nativity group. This project has not only sharpened my ability to consider economic scenarios in real life, but it has given me skills in Python and data analysis that are indispensable in research and data science.

My work over the Ford period focused on understanding the mechanism described in a paper by Chaumont and Shi, 2022, by replicating it using python. Initially, it was challenging for me to sufficiently understand Chaumont and Shi's work without much prior knowledge of the field of job transitioning and wealth study. With the aid of Professor Argudo who patiently answered all my questions over several meetings, I grasped enough to start my own replication. The process was still demanding, but through trials and errors, I have acquired similar results using the method of Value function iteration (VFI). I am familiar with VFI thanks to ECON 304 with Professor Argudo from the previous semester, but tackling such a task with so much independence as I have in the Ford program was something I have never done before. For me, It was a pleasure to receive this opportunity through Ford to assist in professor Argudo's research and challenge myself outside of my comfort zone.

This summer, we worked with Professor Frye and Professor Kagy to research the economic consequences of childhood exposure to environmental toxins. We primarily focused on evaluating whether early exposure to lead would cause differences in intergenerational mobility, occupational rank, and total years of schooling.

The project involved constructing a highly intricate data set that pulled together information on people and towns in the northeast, where the integration of central water systems was prevalent around the turn of the twentieth century and the use of lead and non-lead materials for the municipal pipes varied. We used IPUMS to collect census data for the late nineteenth and early twentieth centuries, the Census Linking Project to probabilistically match people across censuses, the Census Place Project to identify where these people lived, and two books that organized information on cities’ water systems and water chemistry.

Due to the large sizes of the data, we relied on Amazon’s Web Server to run code in RStudio that could compile and build our data set. Our final data set allowed us to identify children in the Northeast as living in either lead or non-lead cities, and then track this cohort overtime to evaluate their outcomes in adulthood.

Exploratory data analysis, involving graphs, maps and summary statistics tables, shows early implications that exposure to lead in childhood does have negative long-term economic consequences. For example, the average income in 1940 for a child exposed to lead in 1900 was $6062 less than the average for a child who was not exposed to lead via city service pipes.

This project also sparked an interest in current-day exposure to toxins and made us more cautious about what we may be consuming in our day-to-day lives. Moreover, this project was a very rewarding experience that enabled us to greatly improve our coding skills and collaborate with our professors, who offered deep insights into working in research. We are grateful that we had this opportunity to learn first-hand about economic research and be part of an exciting and innovative project.

This summer, we worked with Professor Ge and Jiebei Luo on the U.S. Airline project. Specifically, we aim to discuss the product quality and consumer response using the tweets that we scraped from the Twitter API. In the case of the airline industry, the number of airlines in a competitive market is low due to scale economies. Some airlines maintained their monopoly status on specific routes, and their failure to improve service quality (worsening delays) has raised nation-wise concerns and increased consumer dissatisfaction. As social media platforms (e.g., Twitter in our case) provide a large amount of text data, we are able to locate tweets directed at a specific airline through mentions (@). See 1-1 below for sample tweets. Considering the short span of the program and the exploratory nature of our project, we limited our data coverage and focused on two chosen airlines, Alaska Air and JetBlue (Delta and Southwest Air in Madhav’s case), on a sample date pre-covid (Nov 12, 2019).

1-1 Sample tweets

To retrieve historical tweets, we utilized the Twitter API, a platform that allows you to find and retrieve, engage with, or create a variety of different resources including the following: user ID, time created, text, and more. Example 1-2 lists the tweets we pulled. Since many tweets contain URLs, digits, and other things that might disrupt the sentiment analysis, we preprocessed the tweets to trim them and make them ready for future steps.

The purpose of preprocessing is to get rid of any unwanted or irrelevant text elements. To conduct sentiment analysis, we used two different toolkits, TextBlob and Vander, in order to compare between results. However, computer programs have problems recognizing things like sarcasm and irony, negation, jokes, and exaggeration–things that are easy for a human to sense and identify. Failure to recognize these things can skew the results. And since some tweets contain multiple mentions, even when the overall sentiment is negative, the sentiment toward a specific airline is not necessarily negative. Therefore, after putting the texts through automated analyzing programs, we manually identified each of the tweets. See 1-3 for an example when sarcasm cannot be detected by the programs.

Our next step would be to create visualizations to help illustrate our analysis. This summer has been really rewarding as we had no experience in scraping data, preprocessing data, and text/sentiment analysis prior to this project. Exploring new areas is always exciting and learning about the airline industry is relevant to our day-to-day life. Even though the research has just started, and we still have a long way to go before we draw any conclusion, being able to start from scratch and building new skills and knowledge on every step of the way is beyond precious. We would like to thank Professor Ge and Jiebei for this wonderful opportunity and experience.

Prior work classified all non-profits that have publicly available IRS data as either Democrat leaning or Republican-leaning by using a native Bayesian classifier to compare the frequency of the words used by the nonprofit's mission statement with the frequency of words used in speeches given by members of Congress. However, the classification could be made more accurate by neural networks which take advantage of the structure of the sentences used, not just word frequency. We implemented convolutional neural networks and recurrent neural network models that were trained using Congressional data in order to better classify the non-profits.

The code was implemented using python, with congressional data obtained from the Stanford SSDS and DW NOMINATE scores:

This summer, Professor Ho and I worked on a behavioral economics project examining the effectiveness of various apology types on the ability of celebrities and public figures to return to society after being “canceled.”

For our work, we looked first at well-publicized celebrity apologies. We collected a sample of celebrity apologies based on two online databases: a New York Times dataset of celebrities who were publicly reprimanded during the #metoo movement and a dataset compiled by the online apology blog SorryWatch which noted and analyzed a number of celebrity apologies in the recent decades, culminating in over 200 apologies. We then used Amazon Mturk to categorize and assess the effectiveness of these apologies through a survey we constructed.

We additionally worked with academic access from Twitter to download databases of tweets referencing the celebrities within our dataset. These databases of tweets consist of a sample of tweets from six distinct periods of interest:

1. Random collection of tweets from the year prior to the cancellation
2. A database of tweets directly before the news of the cancellable event was published
3. A database of tweets directly after the news of the cancellable event was published
4. A database of tweets directly before the celebrity apologized
5. A database of tweets directly after the celebrity apologized
6. Random collection of tweets from the year following the issuing of the apology

We then used the open source program VADERsentiment to analyze the overall sentiment of the tweets contained in our databases to quantify how public sentiment towards the celebrity changed over time. Our initial results are promising and in line with some of our initial hypotheses, and I hope that our continued collaboration on this project will yield interesting results.

This summer, I worked with Professor Kagy on developing an economics seminar on inequality and discrimination. The project consisted of a few primary tasks: find academic and popular press publications and media sources for the class, create class materials for those sources, and build out the applied econometrics that will be taught alongside the class materials.

Inequality and discrimination, in the US at least, have a long, complicated, and sad history tied to status, subversion, and power. While many are fortunate, far too many, especially people of color, members of the LGBTQIA+ community, women, immigrants, etc., have lived, worked, and struggled in systems of power rigged against them. For example, we can look back at the early to mid-20th century redlining policies that froze people of color out of the housing market. Since assets and wealth, such as a home, build wealth and intergenerational wealth, the impacts of redlining are still being felt today (see Netflix Explained racial-wealth gap & billionaires).

Moving beyond the history and literature of inequality and discrimination, much of my time was spent working with US Census data to create visualizations of inequality over the last 50 years.

The census data I worked with focused on income, and I was able to visualize numerous examples of inequality. I found it immensely striking how evident the inequality was, even after controlling for exogenous variables. As someone who wants to work in the development/general helping people fields, this project has given me further impetus to pursue this path. In addition, after being abroad, working on this project provided a much-needed refresher on coding and econometrics.

I hope my contribution to this project helps whoever takes this class next spring understand and appreciate the complexities of inequality and discrimination, and perhaps push them to do more to help.

Over the past summer, I focused on migration patterns, labor, and marriage markets in Mexico. I also learned data analysis and research techniques with large databases. These databases were used to answer one question: Did men's migration, or lack thereof, after the 2008 recession affect women's marriage markets and labor markets in Mexico? The four databases— ENADID, MCAS, Mexican Census, and ACS— enabled us to make a reliable conclusion about Mexican migration. With that information, we could determine where in the United States Mexican migration happens as well as from where in Mexico Mexican migration starts. The MCAS database is conclusively made up of the Matrículas Consulares issued, a document issued by the Mexican government that provides outside countries, such as the US, proof of Mexican nationality. After acknowledging the similarity with the other databases, using MCAS, we could determine whether migration of working age Mexicans decreased or increased after the recession. We compiled maps of Mexico, including data from each source (ENADID, MCAS, etc.) to make conclusions about where Mexican migration starts and determine the highest sending states. We compiled many graphs not included in the presentation, such as average women's schooling per year, average men's age per year, and marriages in municipalities based on high/low migration states, to name a few. These graphs were created to understand how marriages were affected because of the change in migration to the United States. All of the above helped us gather assumptions about women's marriage and labor markets in Mexico. Although I didn't necessarily get to answer the question because of the many embedded factors, I can't wait to continue working on the project this coming semester!

Cyberattacks are becoming a larger part of what all industries, not just financial firms, have to deal with. Cyber risk exposure has an economically and statistically significant negative effect on the stock market performance of affected firms. Additionally, there is evidence of contagion effects: idiosyncratic firm-level cyber risk has the potential of spreading through interconnected financial markets (same country and industry). In 2011, the Securities Exchange Commission (SEC) released its initial cybersecurity guidance after a spike in cyber attacks, encouraging companies to pay more attention to cyber risks. With Professor Savaser, we are working on a project that aims to document and examine the US bank holding companies’ governance strategies to mitigate cyberattacks. The goal is to construct a measure that captures banks’ cybersecurity proactiveness and examine the type of governance structures that are most effective in dealing with a cyber attack.

We approach the question in multiple stages. First, using text analysis, we document how often cyber terms are mentioned in banks’ annual reports and proxy statements. Second, we analyze the context in which the cyber terms were used. Third, we identify which banks mention cyber risk factors in their disclosures before they or their peers experience an attack or before the SEC published its first guidance of the matter, and which banks act retroactively. Lastly, we merge the dataset with the publicly available data on cyber attacks to examine the relationship between the type of governance mechanisms utilized by banks and the frequency of cyberattacks they experience. In the future, we aim to further investigate the topic in the context of non-financial firms.

This summer Professor Erin McCloskey and I learned about rehumanization through storytelling. Done remotely, we jumped eagerly into this budding research project, fascinated yet inquisitive of each concept. The intention was to recognize the impact of exchanging personal narratives as a transformative pedagogical approach and its role in strengthening relationships within the triad of home, school, and community through its ability to reconstruct barriers surrounding social identities.

Furthermore, the inclusion of this method within classrooms involving those incarcerated (or formerly), and local university students. Our time spent emphasized the value of self-authorship as a form of resistance against dominant power structures and as a tool for forming authentic and constructivist learning environments with students at the core of the curriculum.

From the beginning, there was a need to gather research from all forms of media such as books, peer-reviewed articles, TedTalks, and documentaries to represent the multimodal storytelling that would take place. We then developed an annotated bibliography that would factor in creating an applicable framework that could be sent to the Institutional Review Board. As the project was in its early stages, we were able to discuss how to create positions of strength within the classroom for different types of narratives. While also discussing the possible drawbacks of confining this approach to an academic setting and addressing the dangers of exchanging traumatic stories. Ultimately, this call to exchange allows for reinterpretations and inclusive solutions that prompt community engagement, activism, and identity-making.

As Professor McCloskey continues her work, I thank her for allowing me to contribute over the summer and acquaint myself with a topic I now harbor a passion for.

Known as the leading advocate for critical pedagogy, Paulo Freire was a Brazilian author, philosopher, and educator who championed education in the fight for liberation, equity, and peace. Through close readings of Freire’s work,

explorations of multimedia analysis of his concepts, and dialogues (both academic and community-based) with practitioners of Frerian pedagogy, Professor Hantzopoulos and I inquired: How have Freire’s theories been implemented into practice in educational settings?

Freire’s Pedagogy of the Oppressed (1970) is a seminal, foundational, and revolutionary text from which scholars, activists, and educators draw techniques and theories to institute and sustain social change. With this text as a basis, Professor Hantzopoulos curated a shortlist of books as a catalyst point for further exploration: Education for Critical Consciousness and Pedagogy of Freedom: Ethics, Democracy, and Civic Courage both by Freire, Teaching Community: A Pedagogy of Hope and Teaching Critical Thinking: Practical Wisdom both by bell hooks.

These books guided me in creating an annotated bibliography of fifty sources ranging from teaching guides, lesson plans and activities, journal articles, book chapters, research databases, videos and documentaries, to podcasts with a focus on the implementation of Freirian theory in educational settings. Each source either contextualized, criticized, or built upon Freire’s ideas. This open-ended exploration allowed me to not only track the ways in which people have employed Freirian practice in everyday life but also track the evolution of Freire’s ideas as they were put through an iterative process of trial, error, and revision in varying contexts.

With this complex and rich bevy of sources, I created an interactive website that chronicles this iterative process; this process revealed to me that Freire’s theories and praxis holds a cyclical paradigm. Paulo Freire’s teachings are still relevant today because they emphasize the powerful influence education has in shaping citizens and society.