This is an interesting finding. I knew that vocabulary was loosely correlated with age, but it looks like the gains are made specifically in one's twenties. People are probably expanding their vocabularies as they are required to learn less common words in college and at the workplace. Also, this finding reminds us that cohort, period, and age effects complicate the study of the Flynn effect.
White Americans born in the 40s and 50s end up with Wordsum averages that are significantly higher than that of the 1930s birth cohort. Those born in the 1960s, however, end up with a mean not significantly different than for those born during the Depression.
Something happened to people born in the 40s and 50s--the Baby Boomers, roughly--that gave them larger vocabularies than the cohorts before or after them. The improvement over the 30s cohort seems understandable, but why the decline in the 60s cohort? The 70s cohorts looks like it will end up the same as those born in the 60s. It seems like something has slipped among whites in the past few decades. Explanations like nutrition don't seem to work. I'm not sure how family change would make a big difference. Declining educational standards? Less reading?