The Congressional Budget Office just released a new paper on R&D and productivity growth. In particular, they were interested in whether it made sense to add R&D to macroeconomic forecasting models. Their conclusions:
unless one is willing to assume large spillovers, the inclusion of R&D in macroeconomic models will have little effect on economic forecasts or policy simulations. Moreover, its inclusion in models does not increase understanding of the historical behavior of productivity.....Finally, it is not clear that adding R&D to macroeconomic models will improve their forecasts or projections...In short, the benefits of adding R&D to macroeconomic models probably do not exceed the costs of doing so.
I don't know about you, but I'm very uncomfortable with this conclusion. I'd like to know whether R&D is any more or less significant in terms of projecting future growth than, say, the budget deficit.
Remember--the answers you get depend on the questions you ask.