This is an exploratory exercise which attempts to identify the potential for carrying out online socio-economic surveys in Bangladesh, taking the question of ‘educated unemployment’ as a test case. The topic is of great interest not just in Bangladesh but also throughout South Asia and beyond, where the issue is of particular concern in the context of rapid growth and rising aspirations amongst young people. Most studies depend either on own data generation or

periodic national level surveys like the Labour Force Surveys (LFS). The former consist of small datasets while the latter have a limited number of relevant variables available for analysis. Thus, easier access to larger datasets with better coverage of variables would be a highly welcome additional resource for researchers and policy makers. It was in fact, possible to rapidly generate a large volume of data using an online platform (Facebook) for this exercise. The data validation approach used here is to compare findings with those reported in the wider literature. Generally, the results obtained from the online survey appear both reasonable and defensible. The estimates of educated unemployment are consistent with other available estimates. The relationship of unemployment to education, gender and location are similar to those reported in the literature. The effect of ‘control’ variables like family size, age and family income, were as expected. In particular, family income (reflecting family influence) emerged as a powerful predictor. The study was also able to throw light on two other aspects of the labour market, including duration of unemployment and salary levels.