In times like now, it is easy to forget that job seekers are not only the unemployed, but also the currently employed. On-the-job search is rather common but difficult to measure, as it is often not openly conducted. But it is important to understand it for policy, for example to evaluate the impact of job creation programs.
Simonetta Longhi and Mark Taylor just finished a couple of papers that offer some insights about these two types of job seekers in the United Kingdom. In the first one, they compare their characteristics and search behavior and conclude that employed and unemployed job seekers are different. My reading is that the unemployed are often stuck in a sequence of low paying jobs and drift in and out of employment. The employed job seekers are rather on the way up and improve their situation at each change. Thus these two types are not substitutes and do not compete for the same jobs.
In the second paper, they compared the job finding probability of both types. Consistently for search theory, they observe that on-the-job seekers take much longer to find jobs, reflecting that they can afford to wait for the perfect match. The unemployed, however, have content themselves with the first offer, especially since unemployment insurance criteria have tightened in the UK. The presence of the former has no impact on the search success of the latter, confirming that they are not substitutes.