408 G. S. Hansen and B. Wernerfelt Null model F Ratio =4.232 F Ratio=17.319 (R=,185R=,141) (R=378R=.356) FRatio=17. 276 FRi P=.007 egrated Model 2 Figure 2. Testing the differences of the three models Table 4. Estimated variance decomposition(percent- assets(Itami and roehl, 1987). If this is true age of variance explained) our estimate of the importance of organizational Economic model factors is biased downward. So our results may Organizational Model 37.78 be even stronger than it first appears. Second one might suspect that high performance allows Error rational mana uld not invest in this were it not for ffect of climate Total 100.00 Third, the empirical results of Denison (1982) found that profitability lagged organizational ntal contributions to R contra adicts this, Fourth most theory in the area postulate this direction of causality. Finally, early and however, several our studies point in the same direction. Nevertheles interpretation. First, it may be argued thatgood our result is subject to the usual discussion of organizational practices help a firm select good causality versus association. The answer depends economic environments, or obtain relative advan- substantially on the model in which one chooses tage through the creation of intangible or invisible to interpret the results