Search
Program Calendar
Browse By Day
Browse By Person
Browse By Theme Area
Browse By Session Type
Search Tips
Conference Blog
Personal Schedule
Sign In
X (Twitter)
Although philanthropy pursues a noble goal, it is in no short supply of criticism. According to Cunningham (2020), philanthropy “was criticised as much as it was praised” (p. 1). This duality of criticism and praise is still very much alive today, with scholars both criticizing (Reich, 2018) and defending (Breeze, 2021) philanthropy. Given the ongoing criticism of philanthropy, paired with recent philanthropy-related scandals (e.g., “The failed philanthropy of Sam Bankman-Fried”, The Washington Post, 2023), the question arises of whether philanthropy faces an image crisis. To answer this question, this work analyzes whether philanthropy appears in a positive or negative light in news articles.
Using the historical U.S. newspaper dataset provided by the Library of Congress’s Chronicling America project allows me to go back as far as 1840 to answer this question. This historical part of the analysis draws on 20 million historical newspaper scans and their corresponding digitized texts. 14,600 of these newspaper articles published between 1840 and 1960 mention philanthropy at least once. This data source is complemented by 5,500 more recent philanthropy mentioning New York Times articles published between 1980 and 2023.
Analyzing whether philanthropy is mentioned in a positive or negative light in a given newspaper article is complex and goes beyond the capabilities of common sentiment analysis algorithms. This is because philanthropy is often mentioned in a negative context (e.g., a natural disaster) but in a good light (e.g., providing help for the victims). Because most existing sentiment analysis algorithms cannot disentangle this, I use GPT-3.5 to analyze whether philanthropy appears in a good or bad light in a given newspaper article. GPT-3.5 was proven very effective for such a task and outperforms existing sentiment analysis algorithms by a large margin (Kheiri & Karimi, 2023; Rathje et al., 2023).
The results show that philanthropy is mentioned in an increasingly positive light from 1840 to 1960. This positive trend continues in the New York Times articles until it reaches a plateau in the year 2005, from where it declines back to the level of 1980. I will use topic modeling to explore the topics in which philanthropy is mentioned negatively and positively, respectively. The results provide new insights into the public perception of philanthropy, complementing recent survey-based approaches (IU Lilly Family School of Philanthropy, 2023).
Breeze, B. (2021). In defense of philanthropy. Agenda Publishing.
Cunningham, H. (2020). The reputation of philanthropy since 1750: Britain and beyond. In The reputation of philanthropy since 1750. Manchester University Press.
IU Lilly Family School of Philanthropy. (2023, April 6). What Americans think about philanthropy and nonprofits. https://scholarworks.iupui.edu/server/api/core/bitstreams/b5904a8a-5081-42cd-bd44-56740b98fb67/content
Kheiri, K., & Karimi, H. (2023). Sentimentgpt: Exploiting gpt for advanced sentiment analysis and its departure from current machine learning. arXiv preprint arXiv:2307.10234.
Rathje, S., Mirea, D. M., Sucholutsky, I., Marjieh, R., Robertson, C., & Van Bavel, J. J. (2023). GPT is an effective tool for multilingual psychological text analysis.
Reich, R. (2020). Just giving: Why philanthropy is failing democracy and how it can do better. Princeton University Press.
The Washington Post. (2023, October 16). Opinion | the failed philanthropy of Sam Bankman-Fried. https://www.washingtonpost.com/opinions/2023/10/16/sam-bankman-fried-fraud-philanthropy/