My ACM membership came due just recently. In the light of objections to their copyright policy, I seriously considered not renewing in protest. I agree with Panos that the ACM should not seek copyright from authors, unless the authors are actually being paid for their work. I'm particularly struck by Bob Carpenter's experience as someone outside academia being asked to pay $20 per article for access to ACM papers---again, papers that the ACM paid no-one for, neither authors nor reviewers. I have recently been helping (as a return favour) a small local IT firm improve the text analysis features of one of their products, which has primarily involved finding and summarizing robust methods from the academic literature; having to pay for each article would have made the process not only expensive but also much more frustrating. One of the key roles of publicly-funded research is surely to make research findings readily accessible to independent entrepreneurs and innovators; placing fee-walls in front of research publications frustrates this aim.
The proper role of the ACM, as an academic and professional society, is to promote and advance computer science as a research and industrial field, not to maintain bureaucratic stasis through revenue raising activities. I recognize that there is a cost in collating, archiving, and distributing publications, but this cost surely can't be that great, and once it is amortized, I don't accept that it is morally correct for the ACM to continue collecting revenue on research that they did not fund, and papers whose authors they did not pay. Nor am I saying this as a cheap-skate; I not only pay my membership dues, but subscribe to the print versions of ACMTOIS and other ACM publications, even though I could get them free in e-format through my university, in order to support the dissemination work of these journals.
Despite these concerns, I decided after much reflection to renew my ACM membership. Several considerations inclined me in this direction. First, the ACM (and other professional organizations such as ASIST and IEEE) have more reasonable copyright terms than the commercial journal publishers, such as Springer and Elsevier. The ACM allows author publication of pre-prints, and archiving of these at pre-print archives; essentially, this means it is up to the author whether their work is publicly available or not.
Access prices are also much more reasonable for ACM than for commercial publishers. For $US200 (which includes ACM membership), you get full access to the ACM digital library. How reasonable a deal this is depends upon the field you're in; but for a typical IR project, the ACM DL will provide a majority of the literature required. The commercial publishers do not offer full access to individuals under any terms; you're forced to pay the $35 per article fee if you're not inside a University paywall, or $600 to subscribe to each individual journal. And remember, the authors and reviewers don't receive a cent of this money.
Finally, since ACM is a member-run organization, there is the potential in the future for more liberal access policies to be instituted, if the members support it strongly enough (and are prepared to fund it); whereas with the commercial publishers, the only motive is profit maximization. In another decade, it could well be that professional organization work is freely available and widely cited, while the commercial publishers are left trying to screw the maximum return from paywalled and increasingly obscure material.
So, membership renewed. What, then, are the implications for publishing and reviewing? Reflection on these questions has made me strongly averse to publication in commercial journals, which in my field means IPM and the IR journal. I recognize that any work I submit here will be taken entirely beyond my own control, without me being paid for it, and locked up indefinitely from reasonably-priced access by non-academics. This is not only an affront to me; it reduces the chances that my work will actually be used (rather than just cited). Therefore, I should not publish in these venues, when there are alternatives. And if I'm not willing to offer my services as an author to these journals for free, I should not do so either as a reviewer. If other researchers agree with my reasons, then I'd encourage them, too, to boycott these and all other commercial journals---providing there are reasonable alternatives.
The situation with the journals of professional organizations is unsatisfactory, too, but for the reasons given above, it is less unsatisfactory than for the commercial publishers. And in the IR field, there are few true open-access alternatives. The only one I'm aware of is the Journal of Digital Information, and I'm unsure how happy they are to receive technically-heavy material. I'd like to see more open access journals in the field, and I'd be prepared to bias submissions in their favour. In the mean time, though, professional outlets such as the ACM offer an acceptable compromise.