This experiment tested how what is communicated about the impact of automation affects the extent to which individuals are concerned about automation, are interested in re/upskilling to prepare for automation and intend to undertake actions accordingly. To address this question, a 2×2 experiment was conducted. Both the effect of the prospected severity of technological developments (severe threat vs. minor threat) and the prospected timeframe of such developments (short-term vs. long-term) on individuals’ re/upskilling attitudes and behaviours were tested. The experiment was introduced in a vignette about the effect and timeframe of automation, where the text of the vignette was varied experimentally. Following presentation of the vignette, a series of follow-up questions about re/upskilling were asked. Additional questions about the respondent’s current job skills and security were included before the vignette for controls. Only employed respondents were included in this experiment. The two dimensions (effect and timeframe) and a control text were combined to be controlled by one variable. The allocation was made at the household level, with 1/5 of households allocated to each group. The treatment allocation variables is n_ff_techworkw14 on record n_hhsamp_ip (allocation stratified by ff_incentw14, sampleorig, ff_gridmodew13, l_hhmodes):
1 Major severity; long-term
2 Major severity; short-term
3 Minor severity; long-term
4 Minor severity; short-term
5 Control group
The variables affected by this experiment are on record n_indresp_ip:
datawork, algorithmwork, robotwork, losejob, losejoblong, techreplace, techdoes, techopp, techimport, techoutdate, techskills, techtrain, techtime, techchange
