Simultaneous Equations Models Prepared by Vera Tabakova, East Carolina University 11.1 A Supply and Demand Model 11.2 The Reduced Form Equations 11.3 The Failure of Least Squares 11.4 The Identification Problem 11.5 Two-Stage Least Squares Estimation 11.6 An Example of Two-Stage Least Squares Estimation 11.7 Supply and Demand at the Fulton Fish Market Principles of Econometrics, 3rd Edition Slide 11-2 Figure 11.1 Supply and demand equilibrium Principles of Econometrics, 3rd Edition Slide11-3 D em and: Q 1 P 2 X e d S upply: Principles of Econometrics, 3rd Edition Q 1 P e s (11.1) (11.2) Slide11-4 E ( e d ) 0, var( e d ) d E ( e s ) 0, var( e s ) s 2 2 (11.3) cov( e d , e s ) 0 Principles of Econometrics, 3rd Edition Slide11-5 Figure 11.2 Influence diagrams for two regression models Principles of Econometrics, 3rd Edition Slide11-6 Figure 11.3 Influence diagram for a simultaneous equations model Principles of Econometrics, 3rd Edition Slide11-7 1 P es 1 P 2 X ed P 2 1 1 X ed es 1 1 (11.4) 1 X v1 Principles of Econometrics, 3rd Edition Slide11-8 Q 1 P es 2 ed es 1 X es 1 1 1 1 1 2 1 1 X 1e d 1e s (11.5) 1 1 2 X v2 Principles of Econometrics, 3rd Edition Slide11-9 The least squares estimator of parameters in a structural simultaneous equation is biased and inconsistent because of the correlation between the random error and the endogenous variables on the right-hand side of the equation. Principles of Econometrics, 3rd Edition Slide 11-10 In the supply and demand model given by (11.1) and (11.2) the parameters of the demand equation, 1 and 2, cannot be consistently estimated by any estimation method, but the slope of the supply equation, 1, can be consistently estimated. Principles of Econometrics, 3rd Edition Slide 11-11 Figure 11.4 The effect of changing income Principles of Econometrics, 3rd Edition Slide11-12 A Necessary Condition for Identification: In a system of M simultaneous equations, which jointly determine the values of M endogenous variables, at least M–1 variables must be absent from an equation for estimation of its parameters to be possible. When estimation of an equation’s parameters is possible, then the equation is said to be identified, and its parameters can be estimated consistently. If less than M–1 variables are omitted from an equation, then it is said to be unidentified and its parameters can not be consistently estimated. Principles of Econometrics, 3rd Edition Slide 11-13 Remark: The two-stage least squares estimation procedure is developed in Chapter 10 and shown to be an instrumental variables estimator. The number of instrumental variables required for estimation of an equation within a simultaneous equations model is equal to the number of right-hand-side endogenous variables. Consequently, identification requires that the number of excluded exogenous variables in an equation be at least as large as the number of included right-hand-side endogenous variables. This ensures an adequate number of instrumental variables. Principles of Econometrics, 3rd Edition Slide 11-14 P E ( P ) v1 1 X v1 (11.6) Q 1 E P v1 e s 1 E P 1 v1 e s (11.7) 1 E P e* Principles of Econometrics, 3rd Edition Slide 11-15 Pˆ ˆ 1 X Q 1 Pˆ eˆ* Principles of Econometrics, 3rd Edition (11.8) Slide 11-16 Estimating the (11.8) by least squares generates the so-called twostage least squares estimator of β1, which is consistent and asymptotically normal. To summarize, the two stages of the estimation procedure are: Least squares estimation of the reduced form equation for P and the calculation of its predicted value Pˆ . Least squares estimation of the structural equation in which the right- hand side endogenous variable P is replaced by its predicted value Pˆ . Principles of Econometrics, 3rd Edition Slide 11-17 y1 2 y 2 3 y 3 1 x1 2 x 2 e1 1. (11.9) Estimate the parameters of the reduced form equations y 2 12 x1 22 x 2 K 2 xK v2 y 3 13 x1 23 x 2 K 3 x K v3 by least squares and obtain the predicted values. Principles of Econometrics, 3rd Edition Slide 11-18 yˆ 2 ˆ 1 2 x1 ˆ 2 2 x 2 ˆ K 2 x K (11.10) yˆ 3 ˆ 1 3 x1 ˆ 2 3 x 2 Principles of Econometrics, 3rd Edition ˆ K 3 x K Slide 11-19 2. Replace the endogenous variables, y2 and y3, on the right-hand side of the structural (11.9) by their predicted values from (11.10) y1 2 yˆ 2 3 yˆ 3 1 x1 2 x 2 e1 * Estimate the parameters of this equation by least squares. Principles of Econometrics, 3rd Edition Slide 11-20 The 2SLS estimator is a biased estimator, but it is consistent. In large samples the 2SLS estimator is approximately normally distributed. Principles of Econometrics, 3rd Edition Slide 11-21 The variances and covariances of the 2SLS estimator are unknown in small samples, but for large samples we have expressions for them which we can use as approximations. These formulas are built into econometric software packages, which report standard errors, and tvalues, just like an ordinary least squares regression program. Principles of Econometrics, 3rd Edition Slide 11-22 If you obtain 2SLS estimates by applying two least squares regressions using ordinary least squares regression software, the standard errors and t-values reported in the second regression are not correct for the 2SLS estimator. Always use specialized 2SLS or instrumental variables software when obtaining estimates of structural equations. Principles of Econometrics, 3rd Edition Slide 11-23 D em an d : Q i 1 2 Pi 3 P S i 4 D I i e i (11.11) S u p p ly: Q i 1 2 Pi 3 P Fi e i (11.12) d s Principles of Econometrics, 3rd Edition Slide11-24 The rule for identifying an equation is that in a system of M equations at least M 1 variables must be omitted from each equation in order for it to be identified. In the demand equation the variable PF is not included and thus the necessary M 1 = 1 variable is omitted. In the supply equation both PS and DI are absent; more than enough to satisfy the identification condition. Principles of Econometrics, 3rd Edition Slide11-25 Q i 11 21 P S i 31 D I i 41 P Fi v i 1 Pi 12 22 P S i 32 D I i 42 P Fi v i 2 Principles of Econometrics, 3rd Edition Slide11-26 Principles of Econometrics, 3rd Edition Slide11-27 Principles of Econometrics, 3rd Edition Slide11-28 Principles of Econometrics, 3rd Edition Slide11-29 Pˆi ˆ 12 ˆ 22 P S i ˆ 32 D I i ˆ 42 P Fi 32.512 1.708 P S i 7.602 D I i 1.354 P Fi Principles of Econometrics, 3rd Edition Slide11-30 Principles of Econometrics, 3rd Edition Slide11-31 Principles of Econometrics, 3rd Edition Slide11-32 ln QUAN t 1 2 ln PRICE t 3 M ON t 4TUE t 5W ED t 6THU t e t (11.13) ln QUAN t 1 2 ln PRICE t 3 STORM Yt et (11.14) d s Principles of Econometrics, 3rd Edition Slide11-33 The necessary condition for an equation to be identified is that in this system of M = 2 equations, it must be true that at least M – 1 = 1 variable must be omitted from each equation. In the demand equation the weather variable STORMY is omitted, while it does appear in the supply equation. In the supply equation, the four daily dummy variables that are included in the demand equation are omitted. Principles of Econometrics, 3rd Edition Slide11-34 ln QUAN t 11 21 M ON t 31TUE t 41W ED t 51THU t 61 STORM Yt v t 1 (11.15) ln PRICE t 12 22 M ON t 32TUE t 42W ED t 52THU t 62 STORM Yt v t 2 (11.16) Principles of Econometrics, 3rd Edition Slide11-35 Principles of Econometrics, 3rd Edition Slide11-36 Principles of Econometrics, 3rd Edition Slide11-37 To identify the supply curve the daily dummy variables must be jointly significant. This implies that at least one of their coefficients is statistically different from zero, meaning that there is at least one significant shift variable in the demand equation, which permits us to reliably estimate the supply equation. To identify the demand curve the variable STORMY must be statistically significant, meaning that supply has a significant shift variable, so that we can reliably estimate the demand equation. Principles of Econometrics, 3rd Edition Slide11-38 ln PRIC E t ˆ 12 ˆ 22 M O N t ˆ 32TU E t ˆ 42W ED t ˆ 52TH U t ˆ 62 STO RM Yt ln PRICE t ˆ 12 ˆ 62 STO RM Yt Principles of Econometrics, 3rd Edition Slide11-39 Principles of Econometrics, 3rd Edition Slide11-40 endogenous variables exogenous variables Fulton Fish Market identification reduced form equation reduced form errors reduced form parameters simultaneous equations structural parameters two-stage least squares Principles of Econometrics, 3rd Edition Slide 11-41 Principles of Econometrics, 3rd Edition Slide 11-42 co v P , e s E P E P e s E e s E P es [sin ce E e s 0 ] E 1 X v1 e s [su b stitu te fo r P ] ed es E es 1 1 E es 2 1 1 s [sin ce 1 X is ex o g en o u s] (11A.1) [sin ce e d , e s assu m ed u n co rre lated ] 2 1 1 0 Principles of Econometrics, 3rd Edition Slide 11-43 b1 b1 Pi 1 Pi e si 2 P i Pi Q i 2 P i 1 E b1 1 Principles of Econometrics, 3rd Edition (11A.2) e 1 2 si Pi Pi hi e si (11A.3) E hi e si 1 Slide 11-44 P Q 1 P P es 2 E P Q 1 E P 1 b1 Q i Pi / N 2 Pi / N E PQ E P E PQ E P Principles of Econometrics, 3rd Edition 2 2 1 2 E Pe s E P es E P 2 E P es E P 2 s 2 1 1 1 E P 2 1 Slide 11-45