THE COGNITIVE EMBEDDEDNESS OF COMPETITION AND ENTREPRENEURIAL PERFORMANCE IN THE NEW YORK CITY RESTAURANT INDUSTRY

人们对产品类别如何以及在多大程度上影响我们比较、审查、评估和消费哪些产品，以及我们选择从哪些公司购买产品非常感兴趣。目前对企业和产品如何分类的兴趣主要集中在错误分类对企业绩效的不利影响上。相比之下，我们对哪些产品和公司层面的属性定义了一个类别中的成员资格的理解就不那么完善了。探索这些属性对于确定产品和企业的相对可比性以及市场的竞争边界至关重要。我们认为这些问题在食品和饮料行业，这是一个理想的设置，以探讨有关产品类别，竞争和公司业绩的问题。鉴于产品类型和类型内的变体多种多样，以及企业在空间上的接近性，通过分类的可比性不仅决定了消费者如何在不同的选择中进行比较和选择，而且决定了企业如何竞争、与谁竞争，以及如何和在哪里建立新的企业。对食品和饮料行业的关注也有利于加强对女性和少数族裔参与如何影响企业竞争力的动态的理解，因为该行业的女性和少数族裔管理人员比任何其他行业都多。在不完全竞争的市场中，确定竞争的边界及其影响对理论家和经济学家都是一个挑战。不完全竞争市场是指生产者可以在产品和/或服务的一个或多个方面沿着彼此区别的市场。然而，差异化的概念本身并不简单，它依赖于买家、企业、供应商和第三方（如媒体评论家、分析师和监管机构）之间复杂的相似性推断。企业和产品可以用沿着大量的属性来描述，并且可以被视为或多或少相似，这取决于用于比较的属性、这些属性是如何定义的、何时比较以及由谁比较。这种复杂的维度是许多理论和实践问题的根源，这些问题涉及如何划分市场，哪些公司在同一市场中竞争，市场的竞争结构，这种结构对公司结果的影响，甚至不完全竞争和竞争优势的含义。我们认为，不完全竞争的市场是认知嵌入在不同层次的分析。这种嵌入性提出了关于各级市场解释的一致性和连贯性的重要问题，也需要使用不同类型的数据来充分理解竞争结构及其影响。我们建议通过使用食品和饮料行业企业之间竞争的多层次数据，并使用这些数据来解释该行业的创业成功或失败，来解决市场文献中的这一差距。我们的数据由存档数据组成（例如，社区的人口组成、公司的收入、第三方评论）和原始数据（例如，来自业主的访谈和调查数据）。我们将首先绘制该行业的五年竞争结构图，然后使用该结构来预测下一年开始的公司失败的风险。

There is substantial interest in how and to what extent product categories shape which products we compare, review, evaluate, and consume, and which firms we choose to purchase from. Current interest in how firms and products are categorized focuses primarily on the detrimental consequences of miscategorization for firm performance. By comparison, our understanding of which product and firm level attributes define membership in a category is less well developed. Probing these attributes is vital to determining the relative comparability of products and firms, as well as the competitive boundaries of markets. We consider these issues in the food and beverage industry, which is an ideal setting to explore questions about product categories, competition and firm performance. Given the variety of product types and variants within type, as well as the spatial proximity of firms, comparability through categorization shapes not only how consumers compare and choose among different options but also how firms compete and with whom, as well as how and where new firms are established. A focus on the food and beverage industry also has the benefits for enhancing understanding of the dynamics associated with how female and minority participation may affect firm competitiveness, because the industry includes more female and minority managers than any other industry. Identifying the boundaries of competition and their effects in imperfectly competitive markets has challenged both theorists and empiricists. Imperfectly competitive markets are those in which producers can differentiate themselves from one another along one or more dimensions of products and/or services. The notion of differentiation itself, however, is non-trivial and dependent upon complex sets of similarity inferences among buyers, firms, suppliers, and third parties such as media critics, analysts, and regulators. Firms and products can be described along a large number of attributes, and can be viewed as more or less similar depending on the attributes used for comparison, how these are defined, when they are compared, and by whom. This complex dimensionality is at the root of many theoretical and practical questions regarding how a market is demarcated, what firms are competing in the same market, the competitive structure of a market, the effect of this structure on firm outcomes, and even the meaning of imperfect competition and competitive advantage. We posit that imperfectly competitive markets are cognitively embedded across levels of analysis. This embeddedness raises important questions about the consistency and coherence of market interpretations across levels, and also requires the use of different sorts of data to fully understand competitive structures and their effects. We propose to address this gap in the literature on markets by using multi-level data on competition among firms in food and beverage industry, and by using these data to account for entrepreneurial success or failure in this industry. Our data consists of the archival data (e.g., demographic composition of neighborhoods, the revenues of firms, third-party reviews) and primary data (e.g., interview and survey data from owners). We will first map the five-year competitive structure of the industry, and then use this structure to predict the hazard of failure of firms starting the following year.