Abstract: A Machine Learning Analysis of Social Work Job Advertisements in the Post-Pandemic Era (Society for Social Work and Research 27th Annual Conference - Social Work Science and Complex Problems: Battling Inequities + Building Solutions)

All in-person and virtual presentations are in Mountain Standard Time Zone (MST).

SSWR 2023 Poster Gallery: as a registered in-person and virtual attendee, you have access to the virtual Poster Gallery which includes only the posters that elected to present virtually. The rest of the posters are presented in-person in the Poster/Exhibit Hall located in Phoenix A/B, 3rd floor. The access to the Poster Gallery will be available via the virtual conference platform the week of January 9. You will receive an email with instructions how to access the virtual conference platform.

503P A Machine Learning Analysis of Social Work Job Advertisements in the Post-Pandemic Era

Saturday, January 14, 2023
Phoenix C, 3rd Level (Sheraton Phoenix Downtown)
* noted as presenting author
Cheng Ren, MSSA, PhD Student, University of California, Berkeley, Berkeley, CA
Emmeline Chuang, PhD, Associate Professor, University of California Berkeley, Berkeley, CA
Julian Chun-Chung Chow, PhD, Hutto-Patterson Charitable Foundation Professor, University of California, Berkeley, CA
Brenda Mathias, MSSA, PhD Student, University of California, Berkeley, Berkeley, CA

Social work is one of the fastest-growing occupations in the U.S., with 12% projected job growth between 2020-2030 (compared to 8% for all occupations). The COVID-19 pandemic and in many states, recent or planned reforms in the child welfare, health, and behavioral health sectors have resulted in major changes to the nature of social work. Yet, little empirical information exists regarding changing job demand, and/or to the specific qualifications or skills that employers are seeking in their social workers. This study draws on a database of social work job advertisements to address the following questions: 1) What skills and other job qualifications are employers currently seeking in social workers? 2) To what extent do these skills and qualifications vary across fields.


We used Python to build a web scraper to extract all job advertisements with the keyword "social work" available on Indeed.com, the most frequently visited job site in the world, in March 2022. The raw HTML of each post was parsed to identify job titles, salary, organization, location, and job description. Duplicates were eliminated, resulting in a final total of 1,086 unique jobs. We then applied natural language processing techniques (NLP), specifically "BERT", to convert the language to vectors. An unsupervised machine learning technique, cluster analysis, was used to categorize job descriptions by social work field. Named-entity recognition (NER), implemented using the Python Spacy package, was used to identify the most in-demand skills or qualifications in each field.


Jobs were clustered in the following fields: healthcare (48.1%), schools (15.6%), other child and family (13.8%), and behavior therapy (9.2%). Around half (52.7%) of job descriptions mentioned master's degrees (61.3% in healthcare compared to 44.7% in other fields). Over half of job postings (59.2%) also described licensing requirements (e.g., LCSW, LPC, LMFT). Only 5% of posts mentioned telehealth.

Several common top skills are mental health care, case management, and communication. Discharge planning and crisis intervention are particular top skills for social workers in healthcare. In schools, the unique top skills identified included scheduling/planning and social-emotional support.

Conclusions and Implications:

The study identifies healthcare as the largest field in recruiting social workers. Over half of postings were for those with MSWs or an equivalent master’s degree; licensure was also highly desirable. Across different subfields, we can identify common skills (e.g., mental health care) as well as special ones (e.g., discharge planning). These findings provide important information from the market side, which could help schools and students of social work better monitor the environment of their current or even future employment and adjust correspondingly. Furthermore, by collecting historical data like this, we can understand how the demands from each sub-field grow or shrink and explore how the social work field reacts as social issues change. Also, this study shows the possibility of how advanced techniques in data science can efficiently benefit the social work field.