Oskar Morgenstern was i of the founders of mathematical game theory, equally co-author with John von Neumann inward 1944 of Theory of Gam...
Oskar Morgenstern was i of the founders of mathematical game theory, equally co-author with John von Neumann inward 1944 of Theory of Games in addition to Economic Behavior. He was equally good the writer inward 1950 (revised inward 1963) of a majority that should live on of substantial involvement to philosophers of the social sciences. The majority was called On Accuracy of Economic Observations, in addition to it called into inquiry the unreasonable flat of confidence that economists had inward economical statistics inward the post-war period. Here is how Morgenstern described his goals inward the start edition:
This study is the preliminary outcome of a prolonged concern with the properties of economical data. Most lately the job of accuracy arose over again inward connecter with an investigation into the construction of the American economic scheme past times agency of input-output tables. These are arrangements of the sales in addition to purchases of industries which exhibit the interconnectedness of the entire economy. They are attempts to render information Quesnay intended to give inward the ancient "Tableau Economique"; they were start used past times W. W. Leontief. (v)He puts the key message of the majority inward these words inward the preface to the minute edition:
This majority addresses itself to both the full general reader in addition to the professional person economist. The old volition come across that decisions made inward job organization in addition to inward world service are based on information that are known with much less certainty than mostly assumed past times Earth or the government. The minute volition discovery that fifty-fifty most widely accepted figures often remove keep mistake components of unexpected magnitude, in addition to consequently cast dubiety on many currently accepted analyses inward economics. (vii)Morgenstern's analysis inward this majority is a real sober assessment of the challenges that aspect upwards the job of assembling national or regional economical statistics -- employment, production, prices, trade, in addition to hence on. In Part Two of the majority he has divide chapters on trade, prices, mining, agriculture, employment, national income, in addition to growth.
His perspective is a scientific one; he is fully supportive of the quest to care for economical reality on the terra firma of empirical data. But he insists on the crucial importance of carefully reviewing the instruments through which our economical observations are made, in addition to assessing the cumulative attain of mistake that is created past times those instruments. In other words, he constructs an applied epistemology to accompany economical observation: a detailed accounting of the reach in addition to limits of accuracy inward these observations.
Our problems begin, Morgenstern believes, with the fact that economical statistics are mostly reported without whatever indication of mistake estimates. "Yet it ought to live on clear a priori that most economical statistics should non live on stated inward the mode inward which they are unremarkably reported, pretending an accuracy that may live on completely out of attain in addition to for the most business office is non demanded.... It volition live on seen afterward that national income in addition to consumers' spending mightiness in all likelihood cannot live on known similar a shot inward business office without an mistake of +/- 10 to xv percent" (8).
Morgenstern specifically discusses 10 major sources of mistake inward the estimation of economical statistics:
- Lack of designed experiments
- Hiding of information, lies
- The grooming of observers
- Errors from questionnaires
- Mass observations
- Lack of Definition in addition to classification
- Errors of instruments
- The factor of time
- Observations of unique phenomena
- Interdependence in addition to stability of errors
The information are based on a geographically stratified random sample of almost 35,000 households. These households are randomly chosen from sampling cells which remove keep themselves been chosen according to a multi-stage cluster sampling physical care for involving the role of probabilities proportional to size, from over 300 geographical areas. The sample is taken inward the middle of each calendar month to maximize comparability with other monthly series. (219)Each of the 4 bureaus mentioned hither acquit their observations using somewhat dissimilar methods; in addition to each sometimes larn inward at significantly dissimilar estimates. For example, for the yr 1950 the Census guess of non-agricultural job was 52.3%; BLS flora 45.2%; in addition to BES flora 34.3%. "The 3 counts of non-agricultural job differ widely past times 4 to 18 meg people [from 1946 to 1961]" (224). Each is a reasonable in addition to justifiable estimate; but at that topographic point are broad differences with them.
Here is an interesting hypothetical graph that Morgenstern introduces to demonstrate how sensitive some economical judgments are to variations inside the mistake attain of a given variable. The venture graph plots national income against fourth dimension equally observed past times an official national information fourth dimension series. The lightly shaded expanse represents the mistake attain around this variable. And the dashed draw represents a hypothetical graph of the movements of national income that is alone consistent with the empirical information associated with the national guess -- but leads to diametrically dissimilar conclusions almost the timing of the job organization cycle.
None of this should live on taken to imply that nosotros shouldn't pay attending to economical statistics, or that at that topographic point are non rigorous methods for attempting to stair out these kinds of variables. Rather, Morgenstern's betoken is that all scientific observation involves complex instruments; in addition to it is incumbent upon the researcher to carefully assess the sources in addition to magnitude of possible errors inward his or her observations. Further, economists in addition to policy makers quest to practise appropriate humility inward making role of economical statistics to back upwards this policy reform or that; they quest to recognize the possibility that the estimates in addition to fourth dimension serial of statistical variables upon which they rely are significantly skewed in addition to inaccurate.
I wonder whether contemporary econometricians would seriously disagree with Morgenstern's skepticism almost claims of several decimal points of precision for variables similar national income, growth, or employment. Is at that topographic point a current-generation refutation of Morgenstern's concerns?
(G. William Skinner wrote a highly interesting article inward 1987 almost the problems inherent inward population, tax, in addition to production information for historical Sichuan; "Sichuan's Population inward the Nineteenth Century: Lessons from Disaggregated Data" (link). His analysis includes some of the same sources of mistake in addition to bias that Morgenstern identifies.)
(Ian Hacking's The Emergence of Probability: Influenza A virus subtype H5N1 Philosophical Study of Early Ideas almost Probability, Induction in addition to Statistical Inference is a real interesting handling of the history of the evolution of the mathematics of social statistics.)
COMMENTS