I publish a business newsletter with a circulation of about 400 that serves an industry with about 2,000 - 3,000 people in it. There are a couple of additional points that Jon and Chuck haven’t touched on directly. They may relate more to business newsletters than academic journals, but the principles are similar.
One is that the cost of such a publication is a business expense, and thus is tax deductible, which effectively reduces the cost. As JC says, the price is not (always) paid directly by the person using it, but comes from a corporate budget, so its cost is not considered in the same context as personal purchases. A $300 journal stands in about the same relation to a $1 million/year business as a $30 magazine does to someone making $100,000 a year. Petty cash.
Unlike general interest magazines, a journal may be virtually the only source for specialized information that is crucial to its readers. If you don’t want Time you can get more or less the same information from Newsweek or any number of other sources. My subscribers, however, have essentially no other source for the information I provide. Even in industries where several journals compete, there are relatively few sources, compared to mass media. This obviously tends to increase the value and price.
In the case of business newsletters, the fact that the information can save (or earn) the organization money justifies what might appear to be high costs. There are newsletters for industries like oil and telecommunications that cost thousands a year, but are considered a good value by their subscribers.
Finally, since Jonathan didn’t make it explicit, I’ll just point out that printing is generally not a small publication’s greatest expense. I charge my subscribers about $300 for a newsletter that costs me about $1.00 per book to print. So my material costs are $12 per subscriber per year. But I have to travel to industry conferences, pay long distance phone bills, send out marketing materials to get new subscribers (the largest single expense), and keep myself in beer and Big Macs. (It’s basically a one-person operation.) So although printing costs per unit go up on shorter print runs, that has little to do with the final price. The expenditures of the whole operation have to be spread out across the number of subscribers, and when they number in the hundreds or thousands, rather than millions, the subscription price has to go up.
I hope that’s helpful.