## The Error Term in the History of Time Series Econometrics.” Econometric Theory (2001)

Citations: | 4 - 1 self |

### BibTeX

@MISC{Qin01theerror,

author = {Duo Qin and Christopher L. Gilbert},

title = {The Error Term in the History of Time Series Econometrics.” Econometric Theory},

year = {2001}

}

### OpenURL

### Abstract

We argue that many methodological confusions in time-series econometrics may be seen as arising out of ambivalence or confusion about the error terms. Relationships between macroeconomic time series are inexact and, inevitably, the early econometricians found that any estimated relationship would only fit with errors. Slutsky interpreted these errors as shocks that constitute the motive force behind business cycles. Frisch tried to dissect further the errors into two parts: stimuli, which are analogous to shocks, and nuisance aberrations. However, he failed to provide a statistical framework to make this distinction operational. Haavelmo, and subsequent researchers at the Cowles Commission, saw errors in equations as providing the statistical foundations for econometric models, and required that they conform to a priori distributional assumptions specified in structural models of the general equilibrium type, later known as simultaneous-equations models (SEM). Since theoretical models were at that time mostly static, the structural modelling strategy relegated the dynamics in time-series data frequently to nuisance, atheoretical complications. Revival of the shock interpretation in theoretical models came about through the rational expectations movement and development of the VAR (Vector AutoRegression) modelling approach. The so-called LSE (London School of Economics) dynamic specification approach decomposes the dynamics of modelled variable into three parts: short-run shocks, disequilibrium shocks and innovative residuals, with only the first two of these sustaining an economic interpretation.

### Citations

498 | Stochastic Complexity - Rissanen - 1989 |

451 | Econometric Policy Evaluation: A Critique - Lucas - 1976 |

349 | The dynamic effect of aggregate demand and supply disturbances. American Economic Review 79, 655–73. - Palgrave Handbook of Econometrics, Edited by Terence C - Blanchard, Quah - 1989 |

319 | Rational expectations and the theory of price movements - Muth - 1961 |

247 |
Dynamic Econometrics
- Hendry
- 1995
(Show Context)
Citation Context ...model specification and therefore “cannot sustain claims to be a ‘demand’ shock or a ‘monetary innovation’” in the sense that it does not correspond to any autonomously identifiable economic factors (=-=Hendry, 1995-=-, p.61 & p.359). This insight follows from the explicit focus on the marginalisation (variable selection) decision as a crucial component on any modelling exercise. A crude caricature of the LSE metho... |

185 | Rational Expectations, the Optimal Monetary Instrument and the Optimal Monetary Policy Rule - Sargent, Wallace - 1975 |

183 |
Specification Searches: Ad Hoc Inference with Nonexperimental Data
- Leamer
- 1978
(Show Context)
Citation Context ...inding a complete theory” would correspond to eliminating the error term, and that attempts to minimise error variance therefore implied searching for theories which would be as complete as possible (=-=Leamer, 1978-=-, pp.65-66). 8 16The Error Term The specification of u = Rγ , together with the Bayesian approach of imposing a prior density for β , gave Leamer considerable freedom to experiment with various poste... |

157 | Business cycle modeling without pretending to have too much prior theory - Sargent, Sims - 1977 |

144 | The Monetary Dynamics of Hyper-Inflation - Cagan - 1956 |

111 | Formulating and estimating dynamic linear rational expectations models - Hansen, Sargent - 1980 |

93 | The Statistical Implications of a System of Simultaneous Equations - Haavelmo - 1943 |

91 | Investment under uncertainty - Prescott - 1971 |

87 | The Limits to Rational Expectations - Pesaran - 1987 |

73 | On a method of investigating periodicities in disturbed series, with special reference to wolfer’s sunspot numbers - Yule - 1927 |

71 | The Analysis of Stationary Time Series. Almquist and Wicksell - Wold - 1938 |

68 |
Application of Least Squares Regression to Relationships Containing Autocorrelated Error Terms
- Cochrane, Orcutt
- 1949
(Show Context)
Citation Context ...DAE) in a programme called “Analysis of Time Series” (see Gilbert 1989), which resulted in the famous Durbin-Watson test (Durbin and Watson, 1950) and the Cochrane-Orcutt (CORC) estimation procedure (=-=Cochrane and Orcutt 1949-=-). Orcutt was a key figure in the Cambridge programme. His time-series research started with a correlation analysis in relation to the time series used in Tinbergen’s (1939) model. From the time-serie... |

67 |
Economic Forecasts and Policy
- Theil
- 1958
(Show Context)
Citation Context ..., a recommendation very close to that of t t t Tintner nearly ten years earlier (see section 2). The recommendation was adopted by Stone (1954) as well as the Netherlands Central Planning Bureau (see =-=Theil, 1958-=-). The appendage AR(1) of (4.3), which distinguishes between autocorrelated disturbances ut and white-noise shocks ε t in a somewhat similar manner to Frisch’s distinction between stimuli versus aberr... |

56 | Estimation of dynamic labor demand schedules under rational expectations - Sargent - 1978 |

54 | Economic measurements for policy and predictions - Marschak - 1953 |

49 |
The history of econometric ideas
- Morgan
- 1990
(Show Context)
Citation Context ... use of statistical methods, of which least squares regression became the most prominent. The regression error term took on either of two rationalisations: errors-in-variables or errors-in-equations (=-=Morgan, 1989-=-, chapter 7). On the first rationalisation, the error term was regarded as arising purely from imprecise measurement of the theoretical variables, and was therefore devoid of theoretical connotation. ... |

43 | An Asymptotic Theory of Bayesian Inference for Time Series - Phillips, Ploberger - 1996 |

37 |
On the statistical treatment of linear stochastic difference equations
- Mann, Wald
- 1943
(Show Context)
Citation Context ...ach, which was to become the core of the VAR approach in econometrics. 12 7. Innovational Errors: VAR Methodology VAR-type models had been used in econometrics from the days of the Cowles Commission (=-=Mann and Wald, 1943-=-) there was initially no suggestion that these VAR representations were non-structural. When the issues of estimation and identification were 20The Error Term discussed by the Cowles Commission autho... |

36 | The summation of random causes as the source of cyclic processes - Slutsky - 1927 |

33 | Serial correlation as a convenient simplification not a nuisance: a comment on a study of the demand for money by the Bank of England - Hendry, Mizon - 1978 |

32 |
Statistical Confluence Analysis by Means of Complete Regression Systems, Oslo: Universitets 0konomiske Institutt
- Frisch
- 1934
(Show Context)
Citation Context ... see Hendry and Morgan (1989). In confluence analysis, regression residuals were seen as resulting from “disturbances of the data” and were seldom either discussed directly or explicitly represented (=-=Frisch, 1934-=-). In a later attempt at synthesis, Frisch categorised the error terms occurring in both his theoretical and empirical models into two types: “A disturbance is a deviation from that situation which sh... |

32 | The Probability Approach - Haavelmo - 1944 |

31 |
Testing for Serial Correlation
- Durbin, Watson
- 1971
(Show Context)
Citation Context ...his research was undertaken at the Cambridge Department of Applied Economics (DAE) in a programme called “Analysis of Time Series” (see Gilbert 1989), which resulted in the famous Durbin-Watson test (=-=Durbin and Watson, 1950-=-) and the Cochrane-Orcutt (CORC) estimation procedure (Cochrane and Orcutt 1949). Orcutt was a key figure in the Cambridge programme. His time-series research started with a correlation analysis in re... |

29 |
Three econometric methodologies: a critical appraisal
- Pagan
- 1987
(Show Context)
Citation Context ...reasing criticism, and a plethora of rival methodologies emerged from these debates. Two methodologies, the VAR (Vector AutoRegression) approach and the LSE (London School of Economics) approach (see =-=Pagan, 1987-=-, 1995), came into prominence during the 1980’s. Both maintain that presence of residual autocorrelation implies a dynamically misspecified structural model, and that empirical modelling should procee... |

24 |
A Study of the Autoregressive Nature of the Time Series Used for Tinbergen’s Model of the Economic System of the United States
- Orcutt
- 1948
(Show Context)
Citation Context ...e time-series perspective, he viewed the Tinbergen model as a vector autoregressive model and estimated AR(2) (autoregression of order 2) models for each of the 52 economic series Tinbergen had used (=-=Orcutt 1948-=-). These estimates led him to conclude that the series could typically be seen as being generated by the following AR(2) with a unit root (4.1) yi , t = yi , t−1 + 0. 3∆yi , t−1 + ε i , t . Orcutt bel... |

23 | Estimation of parameters in time-series regression models - Durbin - 1960 |

23 | The Econometrics of the General Equilibrium Approach to Business Cycles - Kydlannd, Prescott - 1991 |

20 | Regression Analysis of Economic Time Series, N.V.Haarlem, The Netherlands: De Erven - Koopmans - 1937 |

20 | A Classical Macroeconometric Model for the United States - Sargent - 1976 |

20 |
Linear rational expectations models: A user's guide
- Whiteman
- 1983
(Show Context)
Citation Context ... and autonomy of u t , be it an exogenous or a shock variable. Furthermore, the representation designated e t the “fundamental” role as the driving process of cycles in the endogenous variables, cf. (=-=Whiteman 1983-=-, Chapter 1). 11 However, there is a certain taxonomic ambiguity in assigning to the error term a structural interpretation equivalent to that assigned to specific exogenous variables. For example, Lu... |

17 | The Estimation of Simultaneous Linear Economic Relationships - Koopmans, Hood - 1953 |

14 | The estimation of relationships with autocorrelated residuals by the use of instrumental variables - Sargan - 1959 |

11 | The Measurement of Consumers’ Expenditure and Behaviour - Stone - 1954 |

9 | Propagation Problems and Impulse - Frisch - 1965 |

9 | Statistical versus theoretical relations in economic macrodynamics. Mimeograph dated 17 July 1938, League of Nations Memorandum. Reproduced by University of Oslo in 1948 with Tinbergen’s comments. Contained in Memorandum ‘Autonomy of Economic Relations - Frisch - 1938 |

7 | National Product Since 1869 - Kuznets - 1946 |

6 | Three Econometric Methodologies: an Update - Pagan - 1994 |

6 |
The Formation of Econometrics: A Historical Perspective
- Qin
- 1993
(Show Context)
Citation Context ...” (1944; p89). Haavelmo’s work provided a blueprint for the research programme of the Cowles Commission for Research in Economics, when they set to formalise econometrics in the early 1940s (see e.g. =-=Qin 1993-=-). Indeed, their core achievement was in the identification and estimation of the dynamic SEM of a set of variables { y, z} variables, z a vector of explanatory variables): ’ (3.1) yt + Γwt = t A ε ′ ... |

6 |
Statistical Testing of Business Cycle Theories, Geneva: League of Nations. 51
- Tinbergen
- 1939
(Show Context)
Citation Context ...n correspond to the errors of the second kind (Tintner, 1940b). However, Tintner’s ideas and methods were largely overshadowed by the excitement of the profession over Tinbergen’s epoch-making model (=-=Tinbergen, 1939-=-). 3. Errors in Equations: Koopmans, Haavelmo and the Cowles Commission 8The Error Term Tinbergen’s famous (1939) model moved business cycle theories from the realm of latent ideals to a world in whi... |

6 |
The Variate Difference Method
- Tintner
- 1940
(Show Context)
Citation Context ... only upon the random variations in the individual series, which could be obtained via differencing, and that the residuals from such analyses should then correspond to the errors of the second kind (=-=Tintner, 1940-=-b). However, Tintner’s ideas and methods were largely overshadowed by the excitement of the profession over Tinbergen’s epoch-making model (Tinbergen, 1939). 3. Errors in Equations: Koopmans, Haavelmo... |

6 | Forward exchange market unbiasedness: the case of the Australian dollar since - Phillips, McFarland - 1997 |

5 | The Maximum Likelihood Estimation of Economic Relationships with Autoregressive Residuals - Sargan - 1961 |

5 |
Specification Errors and the Estimation of Economic Relationships
- Theil
- 1957
(Show Context)
Citation Context ...ortions [of the disturbance] are small or even ‘as small as possible’”, but he attempted nevertheless to justify the minimisation as the criterion for choosing between different model specifications (=-=Theil, 1957-=-). Second, Theil included both residual autocorrelation and cross-equation dependence as potential problems to be dealt with in specification analysis (Theil, 1958, Chapter 6), thereby reviving the qu... |

4 | A re-analysis of confluence analysis - Hendry, Morgan - 1989 |

4 | The Econometric Analysis of Calibrated Models - Kim, Pagan - 1995 |

2 | Finding a Satisfactory Empirical Model - Morgan - 1988 |