Yet again I came across an example of seeing a difference between two groups and immediately jumping to the conclusion that the difference is caused by discrimination.
In his post “This Chart Shows Why Unemployment Is Terrifying For Older People”, BusinessInsider Deputy Editor Gus Lubin presents the following chart and draws two conclusions:
- Older workers have higher standards;
- Employers prefer younger workers.
Interestingly, the permalink of his post is entitled “This Chart Gives Definitive Proof of Age Discrimination In America”. Clearly the chart does no such thing.
Here’s the comment I left at BusinessInsider regarding this article:
I am very surprised – and very disappointed – at the level of simplistic thinking expressed in this post. Gus looks at the chart, sees that older individuals are more likely to experience longer periods of unemployment, and jumps to the conclusions that (a) older people have higher standards and (b) employers prefer younger workers. Neither of these conclusions is supported by the data.
Presumably Gus is assuming that the duration of unemployment for older individuals is higher because they’re holding out for higher paying jobs, and therefore have higher standards. Nowhere in this chart is there any information about job level, nature of position sought, number of job offers turned down for salary, level of responsibility / position within the organization, etc.
Even more problematic, however, is the false conclusion that the duration of unemployment for older individuals is longer because of age discrimination. Contrary to the post’s permalink, this chart does NOT “give definitive proof of age discrimination in America.” There are a variety of factors that explain why the duration of unemployment differs by age, as some other readers have commented. Aside from level of position, salary grade, etc., older people may be unemployed for longer periods because of skills mismatch, unwillingness to relocate, etc.
Unfortunately, this is not the only example of someone looking at a data set, noticing a difference between two groups based on a simplistic analysis, and automatically inferring that the difference is caused by discrimination. In some cases, this leap to automatically inferring discrimination is caused by ignorance. In other cases it may be intentional to advance policy positions, personal agendas, or to create sensationalized headlines. Either way, it is irresponsible and does nothing to advance the discussion of (un)employment and equal rights.
Shame on BusinessInsider for publishing this!
What do you think? Is there a general misunderstanding of statistics – and the scientific method in general, or is it just an attempt to get readers and website hits with sensational headlines?