I frequently hear that academic scholarship is less important than writing for the public because only two reviewers and the author ever read the typical academic article. Certain articles in pay-to-play journals probably are subject to that criticism. And articles in very niche journals might also be poorly read. But if you are writing for standard academic journals, with names that people in your field would generally recognize, you are certainly getting readers. I base this on my own experience, so in what follows, I present a completely anecdotal, but I hope not unreasonable argument for how often a typical academic article gets read.

Now, there are a few advantages here on focusing on my own academic articles publications. One advantage is that prior to this year, I have had 16 peer-reviewed articles published, and in various kinds of journals, but mostly in history, law, and economics journals. And I am neither a scholar with no citations nor a well-known scholar, but someone lurking in the middle of the academic vastness.

Below I present a chart with the year of my article publication in the first column, the journal title in the second column, followed by the journal impact factor, and what I am calling the “minimum total downloads”.

It is impossible to know, of course, how many times people actually read your work. An article might be cited only once for every hundred or so times it is read. I probably have only two dozens citations from people who are not myself. Some articles may be downloaded and read multiple times, or copied for a class. And of course just because someone downloaded your article doesn’t mean that they read it. But downloads are indications of interest and may be the best metric we have to understand how many times your articles are read.

The column “minimum total downloads” is the sum of just four sources: (1) the downloads made via the journal webpage, (2) downloads of my articles on SSRN, (3) downloads of my articles from my personal webpage, and (4) downloads on Researchgate. I also know that my articles have been downloaded from academia.edu, but I don’t pay for the site and I can’t access those analytics.

At any rate, of the 16 articles here, only for 9 of them can I find corresponding download statistics on the webpage of the journal in which they were published. Only 3 of these 9 journals actually have pure download data, while six of them merely show the number of “views” that the article has received. To calculate the “minimum total downloads” above, I only included that number of downloads, not views. This is unfortunate, because it appears in some instances at least that others who viewed the article could read the whole thing online without ever downloading it. For what it’s worth, the “views” on these seven articles ranged from 91 to 758, with an average of 289 views.

Downloads on SSRN and on my person website are clearly unique downloads, but only 4 of my articles are posted on SSRN. ResearchGate, meanwhile, has two different categories, one for downloads and one for “reads.” I only counted the downloads and not the reads, which appear to be instances of a person accessing only the abstract for the article. It is worth noting, I think, that an article abstract has some impact. I, for one, often read article abstracts to get a sense of what people are writing about in a certain field, without ever downloading the full article. A high ratio of abstract reads to downloads might indicate however that your article isn’t very interesting.

My very average academic articles, published in a range of journals, have been downloaded on average 216 times. Remove the most read article from this calculation, and they still average 181 minimum downloads. These articles were all published in the last decade, and some of them are very new, so I can expect that some of them have not reached their readership “half-life” quite yet, that is to say, looking back 30 years from now, I would expect that these articles get half of their reads in the next decade. Latter-day reads and citations, I believe, are particularly common for history articles.

I assume that a download equals a read, but this doesn’t mean that a person who downloads it reads every word. For most of these articles, there is good reason to think that the total number of reads is much higher than the minimum number of downloads I posted above. For example, this list does not include reads of the physical copies of the journals, such as when a person in the library reads the actual thing. It doesn’t appear to include all of the number of times the articles were accessed from university library searches. It doesn’t include all of the times that I’ve sent articles to people by request, or in which I’ve applied to a fellowship and a job and someone actually read my work (although that might be zero times). It doesn’t include peer-to-peer sharing of the articles. Some of my articles are on ethnic history, I know that non-academic readers have shared the articles with friends or family.

Obviously, making your articles available on various platforms will increase your readership. Some of my articles received few reads on the journal website, but a large number on SSRN, or few on researchgate, but many downloads through my own website. As one might expect, the articles which were downloaded most from these online personal pages were those which were behind paywalls at the various journal pages.

To conclude, my argument is simply this: a scholar who writes articles in recognizable journals, even lower ranked ones, should expect that each article will be read about 100 times. If you are lucky and good, it might be read 1,000 times. In the top, say 50 history journals, there are probably no articles that after 10 years have not been read at least a few dozen times.