What are some criticisms of econometrics

Classical institutional economics: reply to Jan Schnellenbach's comment

  • 1.

    I actually avoid the term Methodological holismalthough individual evolutionary-institutional economists may have used it. It is misleading and does not (any longer) hit the core of modern theory of institutional emergence in populations and systems. Here there is no longer any sharp contrast between the individual and the collectivity / system or the transition from individual action to its results, in interaction with others, and in a systemic context (game forms, network types) as well as to emergent system structures and system behavior is methodological and theoretical nowadays clearer. Starting with a 2 × 2 interaction, different game types, then populations, network structures and other conditions of action are introduced. Of Methodological individualism (in the sense of ideal "rationality") can then quickly no longer make sense, because inter-actions in the model quickly assume an essential property of reality: They become complex and for individuals only local, learning and experimenting halfway through. So we like to be involved in research strategy rational choice begin, but we quickly end on a large possible one Diversity is only more local and transitory (so different more rational) Behavioral strategies. And “methodological holism”, the suggestion that there is only one holistic level, turns out to be useless, if not wrong.

    Right or wrong, the concept of holism always tends to suggest a purely collective level. And that's how I read Jan Schnellenbach's review: Holism as collectivism. However, this criticism is now in vain in the face of a modern institutional economics that converges with other “heterodoxies” in the direction of the indicated complexity economics. (This often refers to Hayek including: Well, Schnellenbach's reference to Hayek's “pattern predictions”, “pattern prediction” would also be used by institutionalists.) In Schnellenbach's comment, this results in a number of points of criticism that I can agree with, which I therefore no longer feel addressed.

  • 2.

    The argument that the modern behavioral economics but made the "mainstream" modern and open and in the main Having taken up the older "heterodox" questions and thus made the EIÖ obsolete, in my opinion misses the point: Of course, economic research is no longer feasible today without a new interdisciplinarity with psychology, biology, neuroscience and brain science, i.e. as complex -economic behavioral research, as I have also tried to sketch it out. Mere adherence to the traditional neoclassical core would no longer have sufficient reputation in research. And we all benefit from the progress in behavioral economics. What, on the other hand, explicit “heterodox” and others (e.g. G. Gigerenzer) criticize and what I have criticized is the prevailing strategy of the (Re-) interpretation of complex results, e.g. B. in the traditional forms of Benefit and preference functionto the old Rationality model and at the end then the point (path) Equilibrium conception of the (ideal) Market at least as a reference point (benchmark). This brings you back into the historical message of the old neoclassical mainstream, namely the uniqueness of the (ideal) “market economy”, as well as the associated linear everyday thought patterns and simplistic rhetoric.

    Insofar as economists abandon such reinterpretation strategies, in my opinion they leave the realm of the mainstream and converge, consciously or unconsciously, with modern heterodox orientations. W.B. Arthur, A. Kirman, D. Acemoglu or D. Kahneman are likely to belong to this "movement", although, unlike the average "heterodox", they have access to the top "mainstream" journals.

  • 3.

    And what are they heuristics (Schnellenbach) in behavioral economics different from what has been learned (ultimately also interactively) Institutions and the biases other than indicators of the functioning of social institutions? And what remains of the economic behavior model? The traditional "economic behavior model" (Schnellenbach) is then only an attempt to rescue, a fallback line in the face of dramatic changes in the knowledge, conceptions and "world views" of all modern systems sciences, from evolutionary physics to evolutionary biology, psychology, anthropology ... A more appropriate economic behavior model should contain the elements “interdependence structures”, “interactions”, “behavior options”, “network structures”, “(emergent and transitory) institutions”, “emergent system structures”, “(individual, collective and systemic) dynamics and performances” as standard.

  • 4.

    It also doesn't seem (any longer) convincing to me that the simplistic, linear “economic behavior model”, however informed from complex behavioral research, varies and reinterprets, with research heuristic and methodological ones Self-restraint and Discipline to justify. Simplicism should not be confused with self-discipline (etc.). The suspicion that I have expressed is that it is not about research strategy, theoretical and methodological disciplining but about the deduction of an as unambiguous and welfare-economically interpretable ("optimal") as possible Equilibrium of the "market", at least as a normative benchmark. We no longer need such a “discipline”, which is bought with unacceptable simplicity (largely linear causality): The modern complex method set, yes, up to ABM, creates sufficient discipline in research strategy (abuse in favor of arbitrariness is never excluded), and modern systemic physics z. B. has not exactly become a "soft" science. Evolutionary institutional economics and complexity economics do not mean anything goes.

    And the predictability, which Schnellenbach mentions, the old ideal of the point prognosis, as the alleged advantage of a “disciplined”, “self-restricted” “economic behavior model”, is notorious for being mostly exactly wrong - and then always normatively turned to become: So that the ideal positive prognosis can occur, this and that (de-regulation) measures "for the market" have to be planned and implemented by the state ...

  • 5.

    rationality: Of course, rule-based behavior can be the result of “individual rationality” as long as the relationships are simple, static / constant and transparent enough. And again: In terms of research strategy, we may start with simple rationality, the timeless (short-term) maximization motif in a one-time game with Nash-GG; but already a social dilemma (Commons problem), as is well known, i. d. Usually only with repetition (and rather in populations) and then only through the threat of sanctions, corresponding social learning processes and the Acquisition of a long-term calculation and rationality as a behavioral heuristic and rule (institution) to solve. The “continental divide” between “mainstream” and “heterodoxy” is then no longer “rationality” but the “only” rationality vs. that Diversity of the situationally required (and successful) rationalities and strategies. What today is "rational" i.S. from target-oriented and successful, tomorrow will be under a different one Social ecology no longer be rational.

    In any case, we can only connect to a simple rationality in the area of Declaration of emergence social institutions. Even when institutions have become independent due to culturally acquired long-term rationality, actors become rational fools (A. Sen), because they act “irrationally” in the traditional sense in the short term. For example, in ubiquitous social dilemmas, institutions cannot become “rational” in the short term, but rather psychologically habitualized Need to become.

    Since this with high sunk costs connected (long trial-and-error and learning processes), z. For example, new, more rational institutions that are better adapted to a new situation do not arise because the old ones are more cost-effective in the short term for individual decisions than the new ones.

    Here old "heterodox" theoretical acquaintances beckon to us everywhere: real uncertainty, Paradoxes and dilemmas, unintended consequences …

  • 6.

    Schnellenbach asks pointedly what the EIÖ theoretically to offer Has. With their great breadth various behavioral options under various (emergent and evolving) institutional arrangements (Allocation mechanisms), which it can analyze with its theoretical modules, the EIÖ, in my opinion, has the great strategic advantage of generating the suggested “theoretical surprises” as they shape reality: Multiple transitory equilibria, idiosyncratic system dynamics, adverse behaviors and selections, and unintended consequences. With the theoretical possibility of non-optimal, adverse behavior and adverse system dynamics and performance, the EIÖ has had the theoretical tools since Veblen critical analysis and not just a description or a normative correction of what is. At Veblen the keyword is that ceremonial Dimension of institutions.

  • 7.

    But behind this is a longer tradition with the highest research relevance and topicality. The program was already at A. Smith created in the TMS: The ambivalent nature of man As an egoistic individual and a social being in need of (adaptation), a whole spectrum of possible individual behaviors and institutions opens up, which are concretized in terms of time and space. Veblens Instinct theory and trickle-down social mechanism had carried the idea forward. Today we've been talking since S. Kauffmans NK model genetic and more epigenetic Complexity from the enormous variability the physical manifestations (human and fruit fly as two from multiple equilibria, with 50% identical genes) and options for action, and the latest brain research shows us that every single action is a (short-term) equilibrium constellation (from multiple equilibria) of countless information transmissions between groups of Neurons is. Smith and Veblen reloaded!

  • 8.

    When Schnellenbach claims such theoretical orientations and areas of possibility (“emotions”, “other motives”) just like that for the “mainstream”, the question arises as to what else could distinguish between “mainstream” and “heterodoxy” or neoclassical or EIÖ. Do we suddenly all agree? Do the faculties no longer need to fight for vocations? Unfortunately it is not so. But he's right: she Distinguishing criteria have increased significantly postponed, after we inquisitive all more complex have become, and more behavioral, and more empirical. M. E. but they have not defused or even dissolved, but only in the area of policy-relevant and ideological (re-) interpretations complex behavioral findings in favor of the old Everyday rhetoric of the "market" postponed (see above).

  • 9.

    Empirical turn: Indeed, one has Econometrics, which was historically closely linked to mainstream economics, but which was analyzed as arbitrary in terms of data technology and function-technology and thus as causally-theoretically irrelevant and abusive (e.g. Keynes; McCloskey / Ziliak; Nell / Errouaki), in various Developed aspects: Cliometry, big data or causal analytical tests are just a few of the key words. The empirical turn is by no means an exclusive trademark of a modernized “mainstream”, as Schnellenbach argues. With physical statistics (and econometrics) and SNA, system dynamics as well as empirically data-supported AB system simulations, a new (and even more comprehensive) “empirical turn” is common to all modern complexity sciences and thus also to large parts of evolutionary institutional economics.

  • 10.

    Policy relevance and policy advice: Here, too, Schnellenbach comes to a head again. Does EIÖ have anything to offer here? If something is neglected here, remains vague and is not convincing enough, that is of course my responsibility, less that of the EIÖ. I referred to Elsner 2017 as a policy-implications paper and related policy literature.

    The traditional discussion about Market power and Welfare losses by "wrong prices But it seems to me again a bit homely: First, "markets" are quite complex (often hierarchically tiered and often spatially overlapping) Networks with absolutely adverse incentives and mechanismsthat not least become persistent right-skewed centrality distributions to lead. SNA will then determine theirs idiosyncratic dynamics and critical (crisis) transitions, and thus also indications for welfare losses. Linear analysis approaches for welfare losses should also take an early exercise in neoclassical logic more seriously, which has shown that even in neoclassically modeled markets under violated marginal conditions there is considerable complexity: the Second best theory (Lipsey / Lancaster). If the first-best cannot be achieved because marginal conditions have been violated, a desired improvement in the (welfare) situation may require the detour of further violations of marginal conditions. Such non-direct policy paths are left out of the linear textbook discussion of policy implications.

    The policy implications of the older institutionalists like with operationalizations of negotiated economies (J.R. Commons) or operationalizations of the Instrumental Social-Value Principle (MR Tool) seem unexciting today, but in any case they were not naive-linear but process-related and already had an adaptive policy conception in mind, which is now also underlined by complexity economics (U. Witt already dealt with a policy conception in an evolutionary crossover perspective in 2003 deservedly done).

    Naturally, this means less one concrete added value generated for the policy concepts practiced, but rather made it clear to what extent the prevailing political system is based on strong, complexity-based, learning and adaptive policy arrangements, and only through a highly complex, qualified and "strong", i. H. in case of doubt, a public system of actors that is capable of massive intervention should also be replaced.