Knowledge regimes
The way most sciences are taught is to start with very simple models in which many unrealistic assumptions are made so that students can learn the big picture and major forces, and then the assumptions are gradually relaxed so that you end up with very sophisticated models. It is true of physics and economics. That was what the First vs. Second best debate between Dani Rodrik and several other bloggers was about a few months ago, to which I had three responses (here, here, and here).
I find the claim among the so-called Second Best camp to be over-simplified for several reasons. The first is the problem of vulgar second best-ism in which they spot an institutional failure, propose a correcting policy, and assume success without investigating whether there are multiple institutional failures which counteract each other, whether there is a private institutional response to the failure, whether the policy actually corrects the problem, or whether the policy has unintended consequences which give rise to a new institutional failure. I have a whole category for this.
The second is their assumption of a knowable, static set of affairs. This is an assumption that the Econ 101 theory is correct, but that the real-world solution of some master equation for universal efficiency and total spiritual creaminess requires state intervention because of those chewy chunks of degradation known as "institutional (market) failure" [1]. This presumes an optimal state of affairs that we should strive for -- the "correct" allocations of inputs, outputs, numbers and types of goods to be made, and prices. This seems to me to be impossible not only because of the unknowability of the current set of all knowledge, but because of the unknowability of all possible knowledge. Hayek was only half right: Not only is the sum of current human knowledge unknowable to a single person, but the sum of all possible knowledge is unknowable to all persons or groups except for the group which consists of all humans over all time.
Although the book isn't explicitly about this, Dengjian Jin's The Dynamics of Knowledge Regimes illustrates a relatively simple case in cultural comparisons, a single slice through the cone. The book is Jin's explanation of the competitive differences between the US and Japan. He notes that previous explanations of Japan's rise fail to explain the current stasis of that economy. Those explanations approach the problem from neoclassical, revisionist, institutional, cultural, technological, and complexity schools of thought, among which the revisionist and complexity schools might be counted as Second Best approaches, the former noting the importance of industrial policy, and the latter noting issues like path dependency. Jin, on the other hand, focuses less on trade and transaction and more on the way in which each culture creates, stores, transmits, and uses knowledge. Each culture has its distinctive isomorphic regime (to use his phraseology), and the two regimes are nearly mutually exclusive.
In Jin's description, the cultures can be identified along the relationship and identity axes, with Japanese falling more into connectual and contextual while Americans fall more into contractual and individual. In those terms, Williamson's contractual schema have little to do with the Japanese experience and therefore are relegated to the status of a subset of the possible relationship schema. The American knowledge regime both results in and encourages the creation of isolated, modularized, disconnected, universal knowledge, while the Japanese regime results in and encourages the creation of highly contextualized, tacit, specific knowledge. Jin also notes that the relationship between the state and industry tends to fall into the same isomorphic pattern, with Japanese government working very closely with the affected industries and American government working (or appearing to work) in a universalist relationship, i.e. DARPA awards contracts for knowledge creation in a competitive bid process while MITI would work closely with an alliance on a development project. Jin's book explores these ideas in detail and also shows how this produces competitive advantage for each culture in distinct sectors. For example, the American approach results in leadership in sectors such as software and biotechnology where talent and knowledge can be modularized and reconfigured endlessly, while the Japanese approach results in leadership in complex fabrication and assembly such as automobile and opto-electronics.
So whereas Americans work with a system which emphasizes contracting, Japanese work within a system which emphasizes long-term relationship building. Asymmetric knowledge and opportunistic breach of contract are therefore rarely a problem in Japan. On the other hand, network effects certainly are a strong problem for the Japanese while the creative destruction machine that is modern America blows through network effects rapidly (and the process appears to be accelerating). Thus, a problem that worries the second-besters in one culture doesn't even make it on to the radar in the other regime.
Now pull back a little and realize that Jin was only comparing dominant Japanese and US knowledge regimes. What would be the result of a similar study of all cultures? Or of subordinate cultures within the US, Japan, and other dominant culture types? Also, the Japanese emphasize tacit knowledge, some of which is destroyed by the simple act of trying to objectify and communicate it, so it is not even clear that we could understand all of the institutional failures in our own culture that a Japanese would note, and vice versa. What would happen if we were to be able to look at our own institutions not only in terms of Japanese understanding, but of all existing, or of all possible cultures?
Now, having made those observations, I immediately begin wondering about things like,
-------------------------------
[1] pro forma, we ignore failures of the Really Big Institution, The State
I find the claim among the so-called Second Best camp to be over-simplified for several reasons. The first is the problem of vulgar second best-ism in which they spot an institutional failure, propose a correcting policy, and assume success without investigating whether there are multiple institutional failures which counteract each other, whether there is a private institutional response to the failure, whether the policy actually corrects the problem, or whether the policy has unintended consequences which give rise to a new institutional failure. I have a whole category for this.
The second is their assumption of a knowable, static set of affairs. This is an assumption that the Econ 101 theory is correct, but that the real-world solution of some master equation for universal efficiency and total spiritual creaminess requires state intervention because of those chewy chunks of degradation known as "institutional (market) failure" [1]. This presumes an optimal state of affairs that we should strive for -- the "correct" allocations of inputs, outputs, numbers and types of goods to be made, and prices. This seems to me to be impossible not only because of the unknowability of the current set of all knowledge, but because of the unknowability of all possible knowledge. Hayek was only half right: Not only is the sum of current human knowledge unknowable to a single person, but the sum of all possible knowledge is unknowable to all persons or groups except for the group which consists of all humans over all time.
Although the book isn't explicitly about this, Dengjian Jin's The Dynamics of Knowledge Regimes illustrates a relatively simple case in cultural comparisons, a single slice through the cone. The book is Jin's explanation of the competitive differences between the US and Japan. He notes that previous explanations of Japan's rise fail to explain the current stasis of that economy. Those explanations approach the problem from neoclassical, revisionist, institutional, cultural, technological, and complexity schools of thought, among which the revisionist and complexity schools might be counted as Second Best approaches, the former noting the importance of industrial policy, and the latter noting issues like path dependency. Jin, on the other hand, focuses less on trade and transaction and more on the way in which each culture creates, stores, transmits, and uses knowledge. Each culture has its distinctive isomorphic regime (to use his phraseology), and the two regimes are nearly mutually exclusive.
In Jin's description, the cultures can be identified along the relationship and identity axes, with Japanese falling more into connectual and contextual while Americans fall more into contractual and individual. In those terms, Williamson's contractual schema have little to do with the Japanese experience and therefore are relegated to the status of a subset of the possible relationship schema. The American knowledge regime both results in and encourages the creation of isolated, modularized, disconnected, universal knowledge, while the Japanese regime results in and encourages the creation of highly contextualized, tacit, specific knowledge. Jin also notes that the relationship between the state and industry tends to fall into the same isomorphic pattern, with Japanese government working very closely with the affected industries and American government working (or appearing to work) in a universalist relationship, i.e. DARPA awards contracts for knowledge creation in a competitive bid process while MITI would work closely with an alliance on a development project. Jin's book explores these ideas in detail and also shows how this produces competitive advantage for each culture in distinct sectors. For example, the American approach results in leadership in sectors such as software and biotechnology where talent and knowledge can be modularized and reconfigured endlessly, while the Japanese approach results in leadership in complex fabrication and assembly such as automobile and opto-electronics.
So whereas Americans work with a system which emphasizes contracting, Japanese work within a system which emphasizes long-term relationship building. Asymmetric knowledge and opportunistic breach of contract are therefore rarely a problem in Japan. On the other hand, network effects certainly are a strong problem for the Japanese while the creative destruction machine that is modern America blows through network effects rapidly (and the process appears to be accelerating). Thus, a problem that worries the second-besters in one culture doesn't even make it on to the radar in the other regime.
Now pull back a little and realize that Jin was only comparing dominant Japanese and US knowledge regimes. What would be the result of a similar study of all cultures? Or of subordinate cultures within the US, Japan, and other dominant culture types? Also, the Japanese emphasize tacit knowledge, some of which is destroyed by the simple act of trying to objectify and communicate it, so it is not even clear that we could understand all of the institutional failures in our own culture that a Japanese would note, and vice versa. What would happen if we were to be able to look at our own institutions not only in terms of Japanese understanding, but of all existing, or of all possible cultures?
Now, having made those observations, I immediately begin wondering about things like,
- What institutional failures are we failing to note?
- How many failures could there be that have yet to be discovered?
- Are there some failures that cannot be detected or described in terms understandable within our culture?
- Since failures may work in both directions, is the net effect of those underprovision or overprovision of the good or service in question? How can we know?
- Because we aren't aware of these failures -- indeed, because the state's relationship falls into the same patterns -- isn't it likely that attempts to counteract them will only exacerbate a set of underlying, undetected problems?
- Even if it were possible to detect all of the possible failures, is it possible to counteract those features which are (A) a distinguishing feature of our society, and (B) only detectable to someone outside our society, and (C) solvable only through techniques which are not available to our society or our state-society relationship? In other words, some problems are apt to be an undeniable feature of our society, but their solution is unavailable to us unless we fundamentally change our society ... in which case many of the other institutional failures and the corresponding responses will be rendered meaningless while we simultaneously choose a whole new set of institutional failures for which we have neither experience nor remedy. At best, we could go back, but then all we have is a mono- or bi-stable system in which we never completely eliminate institutional failure, but rather trade one type for another.
-------------------------------
[1] pro forma, we ignore failures of the Really Big Institution, The State
Labels: book, culture, failure, organization, vulgar 2nd Best Theory
<< Home