Gringras et al (2008) provide an answer in their examination of how publication rates and citation habits vary over the lifespan. They gathered publication and citation data for >14,000 professors between 2000 and 2007. One possible limitation of their analysis is that they examine data from multiple domains (natural sciences, medicine, social science and humanities) which probably have very different baseline publication rates. But still, this is a large dataset and they find some interesting trends.
In their Figure 2 (doi:10.1371/journal.pone.0004048.g002) they show that, looking at all professors (dashed red line), the mean publication rate rises quite quickly up until the age of 40. Here, the publication rate still increases, but not as quickly as before. The publication rate drops slowly after the age of 50.
They also break their data down into a subset of “active researchers” (solid red line) which is a slightly misleading term. I first thought that this was an elite group of people, but this is not what the measure shows. It represents a subset of the total dataset where all people with zero publications for that year are removed. So I would argue that the upper red line needs to be treated with a little caution.
The paper by Gingras et al (2008) is fascinating. But I wonder whether we can actually use this rate (averaged across many researchers) to draw conclusions about any individual researcher? Raw number of papers published over time really is a very crude measure of productivity. Consistent failure to publish year after year for no good reason is bad, but I imagine many people have the odd year here and there with no publications.
Many people in their dataset had years with zero publications. Their Figure 3b (doi:10.1371/journal.pone.0004048.g001) shows that 60% of professors did not publish in the year that they were 30, and in the most productive years (age 40-60) ~40% of professors of a given age will not publish. In fact, an analysis by Ioannidis et al (2014) shows that ~99% of researchers have at least one year with no publications. To have an unbroken track record of publications each and every year is exceedingly rare.
Is it better to publish many papers each year? This is a different question.
Gingras, Y., Larivière, V., Macaluso, B., & Robitaille, J.-P. (2008). The Effects of Aging on Researchers’ Publication and Citation Patterns. PLoS ONE, 3(12), e4048. doi:10.1371/journal.pone.0004048.
Ioannidis, J. P. A., Boyack, K. W., & Klavans, R. (2014). Estimates of the Continuously Publishing Core in the Scientific Workforce. PLoS ONE, 9(7), e101698. doi:10.1371/journal.pone.0101698