(share link) <a href="https://www.wsj.com/articles/where-you-can-still-find-a-work-from-home-job-57d03262?st=5jqywy27029orzf&reflink=desktopwebshare_permalink" rel="nofollow">https://www.wsj.com/articles/where-you-can-still-find-a-work...</a><p>One of the interesting things about this article is its data sources...<p>> Sources: Job postings trends are based on an analysis of more than 250 million employment listings since 2014 by WFH Map, a group of researchers from Harvard University and the London School of Economics and Political Science, among others. Workplace trends are based on analysis by Scoop Technologies, a software company that tracks policies of about 4,500 companies.<p>And that 'group of researchers' is a link which leads you to <a href="https://wfhmap.com" rel="nofollow">https://wfhmap.com</a> (honestly, the wsj article is a restatement and summarization of this data)<p>Digging into this a bit more, they've got methods (after all, that's <i>really</i> what you want to find when you look for data...) <a href="https://wfhmap.com/algorithm/" rel="nofollow">https://wfhmap.com/algorithm/</a><p>> Our Large Language Model (LLM) is built using the DistilBERT attention-based transformer model. Transformers are a deep-learning method in which every output element is connected to every input element of a text sequence, for example, with weights on each element dynamically calculated as the text is processed.<p>> This LLM is then pre-trained on the entire English-language Wikipedia corpus, which helps the framework interpret the intended meaning of a given document or passage.<p>> We further pre-train this model on roughly one million text sequences drawn from our corpus of new online vacancy postings. This ensures the language model is familiar with the language of job ad text.<p>The data can be downloaded - aggregate data is available without signup, more detailed data with signup or upon request.<p>There's also their working paper ( <a href="https://wfhmap.com/wp-content/uploads/w31007.pdf" rel="nofollow">https://wfhmap.com/wp-content/uploads/w31007.pdf</a> ) - not a light read at 58 pages.<p>The abstract for this paper is:<p>> The pandemic catalyzed an enduring shift to remote work. To measure and characterize this shift, we examine more than 250 million job vacancy postings across five English-speaking countries. Our measurements rely on a state-of-the-art language-processing framework that we fit, test, and refine using 30,000 human classifications. We achieve 99% accuracy in flagging job postings that advertise hybrid or fully remote work, greatly outperforming dictionary methods and also outperforming other machine-learning methods. From 2019 to early 2023, the share of postings that say new employees can work remotely one or more days per week rose more than three-fold in the U.S and by a factor of five or more in Australia, Canada, New Zealand and the U.K. These developments are highly non-uniform across and within cities, industries, occupations, and companies. Even when zooming in on employers in the same industry competing for talent in the same occupations, we find large differences in the share of job postings that explicitly offer remote work.