Light duties weren't entirely altruistic. If people were injured due to employer negligence, then compensation could be mitigated by light duties employment. The prevalence of light duties work could probably be found by early New Earnings Survey analysis, where data from 1974 was published in some detail. But you need to understand the Key Occupations for Statistics and the underlying Classification and Directory of Occupations and Trades... Similar detail to the current US classifications that get used a lot.
Slightly more tractable for data analysis would be early waves of the British Household Panel Study, where a work history was taken and coded up in 1990s standards. For people willing to do survey archaeology, the 1975 National Training Survey (available from the Data Archive) does include work histories and reasons for changing occupation. But, coded before value and variable labels were a thing, and the documentation is scanned...
Light duties weren't entirely altruistic. If people were injured due to employer negligence, then compensation could be mitigated by light duties employment. The prevalence of light duties work could probably be found by early New Earnings Survey analysis, where data from 1974 was published in some detail. But you need to understand the Key Occupations for Statistics and the underlying Classification and Directory of Occupations and Trades... Similar detail to the current US classifications that get used a lot.
Slightly more tractable for data analysis would be early waves of the British Household Panel Study, where a work history was taken and coded up in 1990s standards. For people willing to do survey archaeology, the 1975 National Training Survey (available from the Data Archive) does include work histories and reasons for changing occupation. But, coded before value and variable labels were a thing, and the documentation is scanned...
Really helpful (as ever) Paul, many thanks! I have a couple of leads now, so will see if I can find a way to take this forward..