Strategic Management JournalStrat. Mgmt. J., 20: 889–913 (1999)LEARNING TO PLAN AND PLANNING TO LEARN:RESOLVING THE PLANNING SCHOOL/LEARNINGSCHOOL DEBATEPETER J. BREWS1* and MICHELLE R. HUNT21The Fuqua School of Business, Duke University, Durham, North Carolina, U.S.A.2The Kenan-Flagler Business School, The University of North Carolina, Chapel Hill,North Carolina, U.S.A.This paper resolves the long-standing debate between the two dominant process schools instrategy. Analysis of the planning practices of 656 firms shows that formal planning andincrementalism both form part of ‘good’ strategic planning, especially in unstable environments.Environment neither moderates the need for formal planning nor the direction of theplanning/performance relationship, but does moderate firm planning capabilities and planningflexibility. In unstable environments planning capabilities are far better developed and formalplans more amenable to change. The planning/performance relationship is, however, moderatedby planning duration: at least four years of formal planning are required before externalperformance associations are noted. Copyright  1999 John Wiley & Sons, Ltd.INTRODUCTIONA recent bitter debate between two prominentstrategy academicians considers a question vitalto the theory and practice of strategy: what typesof planning should firms utilize in their strategyformation behaviors? Ansoff, flying the PlanningSchool flag, contends that formal planning isbeneficial in both stable and unstable environ-ments (Ansoff, 1991, 1994) while Mintzberg, art-iculating the Learning School view, favors logicalincrementalism, especially in unstable environ-ments (Mintzberg, 1991; 1994a, 1994b). Thispaper presents a resolution to the debate, andreports on a study investigating whether environ-mental conditions moderate the type of planningfirms employ in their strategy formation activities.The impact of the length of time a planningKey words: formal strategic planning, logicalincrementalism, environmental stability, firm perform-ance*Correspondence to: Professor Peter J. Brews, The FuquaSchool of Business, Duke University, Box 90120, Durham,NC 27708-0120, U.S.A.CCC 0886–9383/99/100889–25 $17.50 Received 10 June 1998Copyright  1999 John Wiley & Sons, Ltd. Final revision received 21 April 1999regime has been employed at a firm on theplanning/performance relationship was alsoexplored, together with other Learning Schoolcritiques of the Planning School. After reviewingprior research, the key constructs utilized in thestudy are presented and discussed. Then, hypoth-eses are presented, followed by the study method-ology and a report of the findings. Once the studylimitations are acknowledged, a discussion of thekey implications of the findings, for research andpractice, concludes the paper.THEORY DEVELOPMENTPrior researchFew issues have attracted more attention in strat-egy research than the relationship between themode of strategic planning adopted by the firmand the economic performance of the firm.Regrettably, decades of planning/performanceresearch have yielded inconsistent findings. Areview of 18 empirical studies testing the effectof formal strategic planning on economic per-formance concluded the link was ‘tenuous’890 P. J. Brews and M. R. Hunt(Pearce, Freeman and Robinson, 1987). A meta-analysis of 21 studies found that the formal stra-tegic planning/performance link was weak, witha correlation of 0.1507 (Boyd, 1991). Morerecently, a meta-analysis of 26 studies concludedthat strategic planning positively influenced firmperformance (Miller and Cardinal, 1994), whilea similar analysis of 14 studies investigating theeffects of planning on small firm financial per-formance concluded that the relationship, thoughsmall, was significant and positive (Schwenk andSchrader, 1993). The inconsistencies in findings,and the weak planning/performance relationshipsobserved have been key in the rejection of formalplanning as the ‘one’ best way to plan(Mintzberg, 1994d).One methodological explanation for the incon-sistencies and perhaps the most serious indictmentof early planning/performance research stemsfrom the poor conceptualizations and measure-ment protocols utilized to operationalize the plan-ning construct (Boyd, 1991). Crude dichotomousor trichotomous classifications of planningbehaviors were employed: comparing formal, longrange planners with non-formal, long range plan-ners (Thune and House, 1970) or comparing non-planners with incomplete planners and completeplanners (Kudla, 1980). Following the inconsis-tent and often counterintuitive findings emergingfrom the first two waves of planning/performanceresearch (Pearce et al., 1987) more sophisticatedGuttman scaling techniques (Guttman, 1944)were employed to measure the planning constructin the so-called ‘third wave’ ofplanning/performance research (Pearce et al.,1987). These more sophisticated methodologies(see for example Wood and LaForge, 1979, 1981;Fredrickson and Mitchell, 1984) have in generalproduced stronger planning/performance relation-ships than the earlier work (Priem, Rasheed andKotulic, 1995).A second substantive explanation for the incon-sistencies relates to the impact of environment onthe type of planning employed by firms. Somestudies (for example Eisenhardt, 1989; Goll andRasheed, 1997; Hart and Banbury, 1994; Millerand Cardinal, 1994; Miller and Friesen, 1983;Priem et al., 1995) found that formal strategymaking processes or planning are positivelyassociated with firm performance in unstable, tur-bulent or dynamic environments. Other studiesconcluded formal strategic planning is best suitedCopyright  1999 John Wiley & Sons, Ltd. Strat. Mgmt. J., 20: 889–913 (1999)to stable environments (Fredrickson, 1984; Fred-rickson and Mitchell, 1984; Fredrickson andIaquinto, 1989; Mintzberg, 1973) but recommendincrementalism for unstable, complex, dynamicenvironments facing high uncertainty, disconti-nuity and/or rapid change. The impact of environ-ment on the type of planning firms should employin particular, and the environment’s effect on theplanning/performance relationship in general thusremains unclear.This study utilized a measurement protocolconsistent with the third wave of planning/ per-formance research, and controlled for environ-mental stability in its design. However, unlikeFredrickson (1984), Fredrickson and Mitchell(1984) and Fredrickson and Iaquinto (1989), stra-tegic decision making activities or