Change in Surname Frequency

I recently got news that I'm to be a great (or is it grand) uncle again. It set me to thinking about the prevalence of my surname and a comment in the most recent issue of the Journal of One-Name Studies that "for many surnames the frequency of occurrence as a percentage of the total population appears to be remarkably stable." That's an opinion that's contested in the Letters to the Editor section of the journal.

Using data from Ancestry the bar chart shows the frequency of occurrence of the surname Reid for each census year from 1841 to 1911 for England and Wales combined.
Even allowing for the anomalies of the Ancestry census figures, as mentioned in this recent blog post, there is a clear trend.

As an independent check on the reality of the increase I used FreeBMD to calculate the percent of births in the census years that were Reids. The FreeBMD data is fairly complete up to 1971.
Again the trend is clear and, considering likely year to year variability of births, consistent with the census trend.

The bottom line is to be very wary of assuming a stable frequency of surname population and  vital events.