TWO decades of refinement of macroeconomic and econometric analysis have contributed only modest sophistication to the role in which investment is cast.' The nexus of investment interactions with other national income aggregates has long been recognized as considerably more complex than can be faithfully expressed by casting them in a simple exogenous role as in the earliest Keynesian models. Accordingly, Samuelson's expression of the multiplier-accelerator interaction became a classic early step toward endowing investment with a less aloof character.2 But again, as many others have pointed out, Samuelson's simple, dichotomous separation of all investment into exogenous public investment and endogenous private investment added only a modest amount of realism to the role of investment. An analytical structure that ties private investment wholly to the accelerator certainly constitutes only a second approximation at best. The object of this study is not to attempt to rationalize the full role of investment, either theoretically or econometrically. Such a tour de force will have to come from more ambitious undertakings. Rather, in an effort to make some modest contribution toward the eventual resolution of the exogenous-endogenous investment problem, we present an empirical analysis confined to certain segments of construction activity. In the process of establishing some tentative statistical estimates of one facet of the construction-investment nexus, some reflected light of a more generalized nature may be shed on the causal role of investment in macro-economic change.