Ilse Mintz, Dating American Growth Cycles (1972)

Ilse Mintz’s empirical research chapter develops a chronology of American growth cycles to supplement traditional business-cycle dating. Its central problem is how to identify economically significant slowdowns when aggregate activity continues to rise. The distinction helps explain conflicting assessments of postwar stability: the disappearance of absolute contractions would not necessarily mean the disappearance of cyclical fluctuations. Mintz states the relationship to established business-cycle research succinctly:

The idea is to supplement it gradually by a similar body of information about growth cycles.

The chapter preserves the emphasis on movements shared across many economic activities while distinguishing relatively high- and low-growth phases from expansions and contractions. This terminology matters because classifications can acquire policy implications beyond their analytical purpose:

Labeling a period as a low-rate phase might be interpreted as a criticism of policy makers and as a recommendation of expansionary policies, but labeling it as a traditional expansion would not be interpreted this way.

Mintz treats dating as a framework for investigation, not an automatic prescription for intervention. She also questions chronologies based exclusively on GNP or estimated output gaps. Quarterly reporting and revisions complicate monthly dating, while estimates of potential output introduce another source of uncertainty:

Thus the reference dates will be affected not only by the uncertainties of the basic GNP data but also by erroneous assumptions about the movements in potential GNP.

Two procedures operationalize the growth-cycle concept. Deviation cycles measure departures from a seventy-five-month moving-average trend. Step cycles divide growth rates into successive periods with relatively high or low averages, without fitting an explicit trend. The essential distinction is between the level of growth and its direction of change: deceleration can occur during a high-growth phase, just as improvement can precede the end of a low-growth phase. Growth-rate maxima and minima therefore cannot simply serve as reference turning points. Mintz checks step boundaries against adjacent turns and uses between-phase variance to assess their placement. Agreement between the procedures supports the chronology, although differences in implicit trends and indistinct turning zones produce discrepancies.

A related objective is reproducibility. Mintz extends the Bry–Boschan turning-point procedure to diffusion indexes, which measure the prevalence of cyclical movements, and composite indexes, which aggregate standardized changes. Her provisional seventeen-indicator selection covers employment, production, incomes, spending, inventories, prices, costs, interest rates, and trade. Tests against classical cycles suggest one-month revisions to four postwar peaks while retaining the trough dates. Computerization does not eliminate judgment in indicator selection, borderline cases, or corrections. Nor does it supply a simple amplitude threshold:

Neither the Bry-Boschan program nor the method of this study specifies amplitude minima since such a criterion is very difficult to introduce.

The four combinations of cycle definition and index construction identify seven growth cycles over 1948–69. Mintz provisionally favors the deviation-cycle composite for its intelligibility, smoothness, and responsiveness to movement amplitudes. Cycles associated with classical recessions have roughly twice the amplitude of the others. High- and low-rate phases average nineteen and seventeen months respectively, yielding cycles of approximately three years. Broad diffusion and agreement across indexes support recognizing mild fluctuations, including the weak Korean-period cycle.

The chronology also changes the evaluation of leading indicators. Their fifteen turns correspond one to one with growth-cycle turns, so warnings that appeared false against recession dates can identify genuine slowdowns. Corresponding fluctuations in money-growth rates generally precede growth-cycle turns. These findings suggest that an exclusively contraction-based chronology can obscure systematic relationships and hinder comparison with economies experiencing few absolute declines.

A postscript extends the evidence through August 1970. April 1969 remains the tentative growth downturn, but Mintz’s indexes do not identify a classical recession within the observations examined; her conclusion concerning the first half of 1970 does not settle its subsequent course. The chapter’s contribution is a provisional, empirically tested framework for distinguishing slower growth from absolute decline while retaining continuity with established business-cycle analysis.