Roundaboutness is a concept featured in Aus trian capital theory. Homely stories about the bare handed catching of fish are a prelude to a discussion of the economy's capital structure. The outputs of some stages of production become inputs to others. Production takes time. The capital structure, broad ly conceived, has a temporal profile?one that can be modified in response to changes in intertemporal consumption preferences and resource constraints. This was the central message of Eugen von Bohm Bawerk (1959). Alfred Marshall, who theorized in terms of the short period and the long period, taught us that most problems in economics stem from the ever-critical time element. I think Marshall was right on time. But I also believe that a healthy understanding of some of those problems?particularly the ones in macroeconomics?is not best facilitated by his sim ple short-period/long-period distinction. Adopting Marshallian methods, John Maynard Keynes dealt with the polar extremes in the quality of expectations, casting serious doubt about the via bility of a market system. In the short run, the time element itself is problem: Short-run expectations faithfully reflect reality. If the level of spending changes, the multiplier process plays itself out in a clockwork sequence of spending and earning, even tually achieving a new circular-flow equilibrium. Long-run expectations, however, are another mat ter. Here, the time element is a problem: These expectations, if you can call them that, are baseless. The future is shrouded in an impenetrable fog of uncertainty, leaving the current level of investment spending to be determined by unruly psychological factors?Keynes's infamous animal spirits. The resultant circular flow will gush and ebb and even on average may not entail enough flow to fully employ the labor force. The circular-flow framework, exercised in both its short-run and long-run modes, seems to me to be exactly the wrong framework for understanding and dealing with the time element in macroeconomics. Identifying the polar cases of no and debilitating doesn't get us any closer to a solution to all those intermediate cases lying between the poles. The tell-tale feature that inevitably characterizes this framework has been recognized in recent years by Robert Solow (1997)?namely the lack of any real coupling (Solow's term) between the short and the long run. In Solow's reckoning, the two runs simply divide our discipline's subject matter into (1) the problem of maintaining full employment of existing resources and (2) the determinants of economic growth. A viable alternative to the Keynesian circular flow framework is the Austrian means-ends frame work. People employ means (investment) to achieve ends (consumption). In a capital-using economy, there is a significant time dimension sep arating means and ends. We realize that some pro duction processes take more time than others. And, thinking in macroeconomic terms, we recognize that production time can increase or decrease for the economy as a whole as market conditions war rant. In the medium run (a term from Solow, 2000), the problem is one of adjusting production decisions to (intertemporal) consumption prefer ences?a problem that Axel Leijonhufvud (1998) urges us to put back on our macroeconomic agenda. The only solution available in a decentralized econ omy is one involving entrepreneurial responses to changing market signals and especially to changes in the rate of interest. The time element here (Bohm-Bawerk's roundaboutness of production