Different institutional logics give people different definitions of rational. Market logic says grow fast; clinical logic says do no harm. When they clash in healthcare, banking, or education, technology alone cannot bridge the gap.
A hospital in Alberta replaced its paper-based referral system with a digital one. The goal was efficiency. Wait times dropped. Administrators celebrated. Then physicians started routing patients around the new system because it did not fit how they made clinical decisions. The technology worked exactly as designed. It solved the wrong logic's problem.
I kept noticing this pattern while reading institutional theory for my comprehensive exams. Organizations adopt the same technology and get wildly different outcomes. Not because the technology varies, but because the institutional logics governing those organizations give people different definitions of what counts as rational action. Friedland and Alford (1991) made this precise by arguing that society is an interinstitutional system, a system of multiple institutional orders, each with a defining logic. The market logic values efficiency, competition, and profit accumulation. The professional logic values expertise, peer evaluation, and adherence to standards. The state logic values regulation, public interest, and compliance. The family logic values care, loyalty, and unconditional giving. The community logic values shared identity and collective good. The corporate logic values hierarchy, performance, and planning. The religious logic values sacredness and transcendence. These are not just different priorities. They are different organizing principles that define what is legitimate, who has authority, and what success looks like.
Thornton and Ocasio (2008) sharpened this into a framework. Institutional logics are "a set of material practices and symbolic constructions which constitutes its organizing principles and which is available to organizations and individuals to elaborate," drawing directly from Friedland and Alford's definition. They are ideal types, heuristic devices for capturing the distinctive features of a phenomenon. As ideal types, logics do not precisely conform to reality, but they serve as yardsticks for comparison. Each logic defines what problems get attention, what solutions are considered, who has authority, and what counts as legitimacy. The same organization, the same technology, and the same person can be pulled by multiple logics at once, and those logics can contradict each other.
This is where most technology adoption explanations stop too early. TAM tells you whether a physician finds an AI diagnostic tool useful and easy to use. UTAUT adds social influence and facilitating conditions. Those models predict adoption intention. But they cannot explain why the same AI tool thrives in a bank and stalls in a hospital. To explain that, you need to know which logic governs what counts as a good decision in each setting.
In banking, market logic dominates. Speed, efficiency, cost reduction, and competitive advantage are the organizing principles. When a bank adopts AI for fraud detection or loan underwriting, the logic of the institution and the logic of the technology align. Faster decisions, lower costs, better risk models. The AI tool reinforces the existing logic rather than challenging it.
In healthcare, professional clinical logic dominates. Patient safety, evidence-based practice, peer evaluation, and the duty to do no harm are the organizing principles. When a hospital adopts AI for diagnostic triage, the logic of the technology, efficiency, speed, algorithmic scale, clashes with the logic of the institution. A radiologist who rejects an AI recommendation is not being irrational. Within clinical logic, that rejection is the rational act. The clinical logic says: verify, consult, prioritize safety over speed. The market logic embedded in the technology says: process more cases faster. These are competing rationalities, not competing preferences.
I think this is the most useful lens for understanding why the same technology thrives in one industry and crashes in another. It is not about the technology. It is about which logic defines legitimacy in that setting.
The Faik, Barrett, and Oborn (2020) paper in MISQ takes this further by connecting institutional logics to IT affordances. They argue that IT affordances can shift the balance of power between competing logics. When a new technology affords mass collaboration, for example, the prior institutional logic of centralized expertise and top-down coordination no longer fits, and a logic of distributed contribution can take hold. But the affordance does not by itself shift the logic. Perception and enactment are required. Organizational context can block either step. The affordance opens a path. Adoption of that path shifts the logic. This is a precise mechanism, not a vague claim about technology and culture. IT creates action possibilities. Which possibilities get enacted depends on which logic holds authority. And which logic holds authority depends on who has been reinforced by prior technology choices and institutional arrangements.
Consider education. In many universities, professional logic governs teaching: expertise, peer review, autonomy, and standards. But market logic has been pushing in through rankings, student satisfaction scores, employability metrics, and the language of the student as customer. When a university adopts a learning analytics platform, the affordances of that platform are not neutral. Dashboards that track student engagement metrics reinforce market logic, because they make efficiency, retention, and measurable outcomes visible. Forums that enable peer critique and open discussion reinforce professional logic, because they make expertise, argumentation, and standards visible. The same platform, different logics, different affordances actualized, different outcomes.
I need to double-check this, but my recollection is that Reay and Hinings (2009) studied exactly this kind of competing-logic dynamics in the Alberta healthcare system. They showed how market logic and professional clinical logic coexisted in tension for decades, producing hybrid practices that blended elements of both rather than one logic simply replacing the other. That finding is important because it complicates the simple narrative that technology replaces one logic with another. What actually happens is messier. Logics get layered. They get combined. They get loosely coupled. People draw from multiple logics in the same sentence.
The connection to digital transformation is direct. Vial (2019) defines enabling digital transformation as changes that lead to changes in value proposition and organizational identity. Wessel et al. (2021) tighten the definition: genuine digital transformation redefines the value proposition and produces a new organizational identity, while IT-enabled transformation merely supports the existing one with better tools. But neither definition addresses which logic governs the transformation. A hospital that digitizes its patient records has done IT-enabled transformation. A hospital that uses digital tools to redefine what counts as quality care, who has authority over clinical decisions, and how success is measured has the potential for genuine transformation, but only if the institutional logic shifts too. And logic shifts are hard because they redistribute authority.
This explains why AI adoption in regulated industries looks completely different from AI adoption in tech companies. It is not a matter of technical readiness or organizational readiness in the generic sense. It is a matter of which logic holds the definition of readiness. In a Silicon Valley startup, market logic says: ship fast, learn from failure, disrupt. In a hospital, clinical logic says: validate thoroughly, do no harm, follow evidence. In a public school, professional logic says: teach well, follow standards, protect the vulnerable. When a technology vendor walks into these settings with the same product and the same pitch, they are selling against different definitions of rationality. The vendor sees adoption resistance. The institution sees logic defense.
The practical implication is that anyone designing or implementing technology in a multi-logic environment needs to name the logics before choosing the intervention. If you are introducing AI into a hospital, you need to know whether your implementation plan reinforces clinical logic, challenges it, or tries to hybridize it with market logic. If you are pushing an enterprise analytics platform into a university, you need to know whether the metrics you are making visible will strengthen professional logic or market logic, and what that means for the people whose authority comes from the logic you are displacing. Naming the logics is not optional background. It is the theoretical work that determines whether the technology lands or bounces.
Institutional theory explains why organizations become similar under pressure. Institutional logics explains why they stay different. DiMaggio and Powell (1983) tell you why every hospital eventually adopts electronic health records. Thornton and Ocasio (2008) tell you why one hospital uses those records to cut costs and another uses them to improve diagnostic accuracy. Same technology, different logic, different outcome. The most important thing a technology implementer can learn from institutional logics is not what the technology does. It is which logic the technology serves, and what happens to the people and practices governed by the logic it disrupts.
About the author
Share
More notes
Related notes