Conclusion
We define entrepreneurship in terms of actions that involve some degree of novelty: entrepreneurs start a business, launch a new product, or transform some organization or production process.
We argue the decisions of entrepreneurs are best understood as sets of bundled and subsidiar y decisions representing one “big picture” path—say, opening a restaurant—versus other “big picture” paths, such as staying in paid employment. These bundled sets will be highly complex because, among other things, these decisions will involve nonpecuniary elements in addition to the typically-modeled pecuniary ones. We then argue that Simon's grand critique of the maximization paradigm in economics applies with even greater force because of the complex, bundled nature of the decisions involved. The complexity of making decisions in such an environment suggests “the” optimal solution will be enormously complex to calculate and this complexity is only exacerbated if we consider entrepreneurs will never have the complete set of all relevant data and unlimited cognitive capacity with which to process any such set of data.To conceptualize entrepreneurial decisions as discontinuous, we rely on Yang and his colleagues' inframarginal modeling of producer-consumers. Yang's inframarginal analysis emphasizes why traditional marginal apparatus is insufficient as the decisions, in this context, are corner solutions. As such, agents are presumed to compare total net benefits across the different choice sets.
While we employ Yang's insights into the inframarginal process, we also have qualms. Among his broad objectives was a focus on efficiency conditions in a general equilibrium framework. As a result, he presumes a level of agent optimization that is even more complex than we find in traditional micro theory. Again, we invoke Simon's critique of optimization assumptions: we take the position that—while we accept producer-consumer entrepreneurs likely compare totals across their decisions as Yang suggests—our agents are highly unlikely to be able to fully optimize over those complex-bundled sets.
The process undoubtedly looks a lot more like Simon's “satisficing” than maximizing.When time is added to complex-bundled decisions, entrepreneurs will find themselves locked into certain courses of action once they begin down a particular path because of the transaction costs of acquiring and liquidating assets and because learning through time will make the individual more productive on their current path than on alternative ones. This process suggests entrepreneurial decision-making involves at least some degree of path dependence. We do not, however, interpret path dependence as implying “inefficiency” because there is nothing in our approach that suggests alternative paths are realistically viable: we cannot “wish away” transaction costs or the change in relative productivities.
The implications of approaching entrepreneurship in this manner are manifold. First, it is considerably less misleading to approach the entrepreneurial environment as a highly complex one that involves multiple bundled decisions than one that assumes that entrepreneurs merely need to calculate readily-apparent input price ratios and marginal products. In other words, the realism of the approach suggests greater relevance to the actual context entrepreneurs will find themselves in. Approaching entrepreneurship as a highly complex, bundled, and nonop- timal process is a far cry from, for example, Kirzner's view that entrepreneurs are somehow more “alert” to a free ten-dollar bill.
We believe our model fits a number of stylized facts better than models that assume agents optimize margins or merely arbitrage price differentials across markets. These include work effort and the “all-or-none” nature of entrepreneurship. As such, the bundled nature of the decisions means the entire venture will remain incomplete (think “0” or “1”) until some critical proportion of the bundled set has been accomplished: it is therefore critical to understand that entrepreneurial activities cannot be left partially completed as is implied by the marginal approach.
We also concluded entrepreneurs are likely to consider more than pecuniary returns in their decisions, and there is substantial evidence entrepreneurs work more and probably sacrifice more, such as family life, than similarly situated salaried employees. In addition, our path-dependent result suggests the hurdle to become an entrepreneur and, in fact, to leave entrepreneurship is higher than traditional approaches admit: we should therefore expect fewer individuals to become entrepreneurs and that they will remain in their ventures longer than many traditional external obser vers would predict.Another key implication of our approach is that failure—or at least something short of ringing entrepreneurial success—is likely to be part- and-parcel of the entrepreneurial process: indeed, from our perspective we should be surprised when ventures succeed at all, given the complexity of the decisions, the sheer number of subordinate decisions that must be made, the required effort and dedication, and the additional complications that real time entails.
Acknowledgements We appreciate the research assistance from Keith Stacy and helpful comments from Amanda Barstow, Demi Fink, Emily Heithcock, Garett Jones, Kanybek Nur-tegin, Luis Pollon, David Prychitko, Jesse Pulliam, Nathaniel Smith, Heath Spong, Sanjay Venugopalan, the editors of this volume, and several participants in the Entrepreneurship and the Market Process workshop. We feel it necessary to add the caveat that we do not implicate our commenters or editors as endorsing the position we set out here as several of them disagreed with us, vigorously, at various turns in the argument.