Since operations research deals with optimization problems in which criteria, alternatives, and other parameters are given, in a sense, the solution to the problem lies in the data, and analysis can provide this.
With systems analysis, there are no givens: objectives are multiple, conflicting, and often incommensurable and irreducible to similar terms; alternatives are not identified; the situation is not understood. When the means and ends are interrelated, the problem is not to quantify relationships but merely to identify them. Yet the impulse to quantify and to make exact is strong.
… The theoretical problems of systems analysis – the inability to deal effectively with political and social factors, the inadequacy of definitions of military worth, and the inapplicability of marginal analysis to many military problems – have important implications when considering the interaction between systems analysis and the political context within which it operates.
That was from Clark Murdock’s excellent 1974 book, Defense Policy Formation: A Comparative Analysis of the McNamara Era. Another way of talking about the difference between operations research and systems analysis is that the former is simply a problem of optimization, where the latter combines problems of optimization, prediction, and uncertainty as to how the problem should be framed.
One of the best courses at MIT last year was System Dynamics. I would call it a subset or maturation of system analysis in which the goal is to model the feedback relationships between policies and actions in complex systems. It allows you to account for human behavior. Our first homework was to model an epidemiological problem! (before COVID) I also remember sketching a feedback diagram in class to show that increasing airworthiness policies that take away authorities in order to drive standardization can lead to less compliance because PMs and engineers will “hide” rather than submit due to the loss of autonomy/control over their programs. It was interesting and thought-provoking! If I ever got my PhD it would be to model the shrinking defense industrial base and test the policies that to change it. Here’s a paper that implies the creator of it, Jay Forestor, invented system dynamics as a reaction to the “elitism” of McNamara’s systems analysis. http://web.mit.edu/esd.83/www/notebook/Systems%20Analysis.doc