The Economics and Sociology of Second-Best Organizations

1940年，罗伯特·默顿提出了一种新的组织理论方法。韦伯将重点放在了正式结构(如工作描述、权威关系和职业阶梯)在设计有效组织中的作用上，而默顿则挑战了对正式结构的关注和对高效绩效的结论。在20世纪50年代和60年代，默顿的学生和他们的同龄人对组织进行了详细的第一手观察，并撰写了描述和分析他们所见的书籍(例如，参见布劳(1955)、克罗泽(1963)、道尔顿(1959)、古尔德纳(1954)和塞尔兹尼克(1949))。这些书证实，许多组织没有表现出韦伯理想类型的“精确度、速度、专家控制、连续性、判断力和最佳投入回报”(默顿，1940：561)，而且在许多组织中，非正式结构(如Quid Proquos、共识和未在组织图中显示的关系)至少与正式结构一样重要。此外，这些书展示了非正式结构的两个角色：在某些情况下，它们颠覆了组织的理性设计；在另一些情况下，它们激发了卓越的组织绩效。在20世纪70年代，组织社会学开始将重点从组织内部的生活转移到组织与环境之间的关系上。今天，这种转变基本上已经完成：人口生态学和制度理论是最强大的两个子领域，这两个领域通常都将企业视为黑箱，甚至网络理论也更多地关注企业之间的网络(同样被视为黑箱)，而不是企业内部的人员网络。在经济学方面，过去25年发生了不同的转变，信息理论、激励理论、博弈理论和契约理论迅速发展。我认为，现在是时候通过开发植根于这些新经济理论的模型来重新审视关于组织内部生活的社会学证据了。(我还想知道社会学家停止研究这些问题的一个原因是否是他们已经用完了取得进展的新方法。现在我们有了一个。)这一建议可能看起来并不新鲜：在过去的五年里，有许多关于组织经济学新兴领域的论文。但这种理论化似乎大多是在经验真空中进行的。特别是，我担心许多组织经济学家除了大学之外，没有接触过任何组织！为了避免这种命运，我与一家机构(花旗银行)建立了深厚的关系。在这项提议中，我寻求对第二种方法的支持：我想阅读组织的其他第一手观察者的描述并建立模型，包括来自20世纪50年代和60年代的组织社会学文献的描述，以及横跨商业和学术文献的较新的描述(如鲍尔(1970)、埃克尔斯(1985)和Gerlach(1991))。我对第一手观察者叙述的有限阅读表明，这种研究策略可能会取得巨大的成果。例如，该提案描述了Crozier(1964)的一个论点，该论点被Milgrom和Roberts(1988)重新发明。我的观点既不是说经济学家随后的这种正式建模没有价值，也不是说非经济学家的非正式论点在被形式化之前没有价值。相反，我的观点只是，任何将经济学文献预测到20年后的人都值得进一步关注。因为我读得还不多，所以我不能对我将要开发的模型说太多。因此，在本提案中，我举了我以前工作的两个例子，以表明拟议的研究既可行又有价值。第一个例子来自《认真对待科斯》，这是我应《管理科学季刊》(组织理论的领先期刊)的要求写的一篇关于(排名第二的)低效组织的文章。第二个例子来自Baker，Gibbons和Murphy(1997)，它将Simon(1951)的关系契约概念与Grossman和Hart(1986)的整合定义为资产所有权相结合。第二个例子表明，正式和非正式组织结构不仅共存，而且相互作用，因此，最佳组织设计涉及选择正式设计，以促进有用的非正式互动。

In 1940, Robert Merton launched a new approach to organization theory. Whereas Weber had focused on the role of formal structures (such as job descriptions, authority relationships, and career ladders) in designing efficient organizations, Merton challenged both the focus on formal structures and the conclusion of efficient performance. In the 1950s and '60s, Merton's students and their contemporaries undertook detailed, first-hand observations of organizations and wrote books describing and analyzing what they saw (see Blau (1955), Crozier (1963), Dalton (1959), Gouldner (1954), and Selznick (1949), for example). These books confirmed that many organizations do not exhibit the `precision, speed, expert control, continuity, discretion, and optimal returns on input` (Merton, 1940: 561) of Weber's ideal type, and that informal structures (such as quid pro quos, shared understandings, and relationships not shown on the organization chart) are at least as important as formal structures in many organizations. Furthermore, these books showed two roles for informal structures: in some cases, they subverted the organization's rational design; in others, they inspired superior organizational performance. In the 1970s, organizational sociology had begun to shift its focus from life inside organizations to the relationship between an organization and its environment. Today, this transformation is essentially complete: two of the strongest sub-fields are population ecology and institutional theory, both of which typically treat the firm as a black box, and even network theory focuses more often on networks between firms (again treated as black boxes) than on networks of people within firms. In economics, the last 25 years saw a different transformation, in which the theories of information, incentives, games, and contracts developed rapidly. I believe it is now time to revisit the sociological evidence on life inside organizations by developing models rooted in these new economic theories. (I also wonder whether one reason sociologists stopped working on these issues was that they ran out of new ways to make progress. Now we have one.) This suggestion may not seem new: in the last five years there have been many papers in the emerging area of organizational economics. But much of this theorizing seems to have taken place in an empirical vacuum. In particular, I fear that many organizational economists have been exposed to no organizations other than universities! In an attempt to avoid this fate myself, I have developed a deep relationship with one organization (Citibank). In this proposal I seek support for a second approach: I would like to read and model the accounts of other first-hand observers of organizations, including those from the organizational sociology literature of the '50s and '60s, as well as more recent accounts that straddle the business and academic literatures (such as Bower (1970), Eccles (1985), and Gerlach (1991)). My limited reading of accounts by first-hand observers suggests that this research strategy may be enormously fruitful. For example, the proposal describes an argument by Crozier (1964) that was reinvented by Milgrom and Roberts (1988). My point is neither that such subsequent formal modeling by economists has no value nor that informal arguments by non-economists have no value until they are formalized. Rather, my point is simply that anyone who anticipates the economics literature by two decades merits further attention. Because I have not yet read much, I cannot say much about the models I will develop. In this proposal, therefore, I give two examples of my prior work to suggest that the proposed research is both feasible and valuable. The first example is drawn from `Taking Coase Seriously,` an essay I wrote about (second-best) inefficient organizations at the request of the Administrative Science Quarterly (the leading journal in organization theory). The second example is drawn from Baker, Gibbons, and Murphy (1997), which combines Simon's (1951) conception of relational contracting with Grossman and Hart's (1986) definition of integration as asset ownership. The second example shows that formal and informal organizational structures not only co-exist but also interact, and hence that optimal organizational design involves choosing the formal design to facilitate useful informal interactions.