I read this report a few weeks ago and its finding was interesting but I am concerned with the interpretation.

The authors (I wish I had a link) look at how age is appreciated in the technology sector. This is a great question since the labour market is aging. The authors use the context of one firm (across a dozen locations) where employees propose a process improvement that is posted on the bulletin board. People for the next week can walk by and vote if they like the improvement. No negative voting is allowed.

The authors find that process improvements suggested by older employees (those less than 35 are considered young) are endorsed 8% less than younger employees and that it is highly significant. The people voting have no idea who suggested what so the effect of age seems real. They control many variables like location of plant, week of the year, and what category the process improvement falls in.

They don't know how many people could have voted, so they cant do a ratio, but that does not bother me too much. What I don't understand is that they report that 80% of the employees are considered young and that 71% of process improvements are suggested by younger employees.

Don't they have to show that the number of endorsements is fewer than what we would expect given the age ratio at the firm and not just that older people get less? The authors don't discuss this so maybe it is not an issue, but is it? If it is wrong, how would you do the analysis to show that the older employees are getting less than we would expect given their proportion at the firm?