Tobit model

From Wikipedia, the free encyclopedia
Jump to: navigation, search

The Tobit model is a statistical model proposed by James Tobin (1958)[1] to describe the relationship between a non-negative dependent variable and an independent variable (or vector) . The term Tobit was derived from Tobin's name by truncating and adding -it by analogy with the probit model.[2] The Tobit model shall not be confused with the truncated regression model, which is in general different and requires a different estimator.[3]

The model supposes that there is a latent (i.e. unobservable) variable . This variable linearly depends on via a parameter (vector) which determines the relationship between the independent variable (or vector) and the latent variable (just as in a linear model). In addition, there is a normally distributed error term to capture random influences on this relationship. The observable variable is defined to be equal to the latent variable whenever the latent variable is above zero and zero otherwise.

where is a latent variable:

Etymology[edit]

When asked why it was called the "Tobit" model, instead of Tobin, James Tobin explained that this term was introduced by Arthur Goldberger, either as a contraction of "Tobin probit", or as a reference to the novel The Caine Mutiny, a novel by Tobin's friend Herman Wouk, in which Tobin makes a cameo as "Mr Tobit". Tobin reports having actually asked Goldberger which it was, and the man refused to say.[4]

Consistency[edit]

If the relationship parameter is estimated by regressing the observed on , the resulting ordinary least squares regression estimator is inconsistent. It will yield a downwards-biased estimate of the slope coefficient and an upward-biased estimate of the intercept. Takeshi Amemiya (1973) has proven that the maximum likelihood estimator suggested by Tobin for this model is consistent.[5]

Interpretation[edit]

The coefficient should not be interpreted as the effect of on , as one would with a linear regression model; this is a common error. Instead, it should be interpreted as the combination of (1) the change in of those above the limit, weighted by the probability of being above the limit; and (2) the change in the probability of being above the limit, weighted by the expected value of if above.[6]

Variations of the Tobit model[edit]

Variations of the Tobit model can be produced by changing where and when censoring occurs. Amemiya (1985, p. 384) classifies these variations into five categories (Tobit type I - Tobit type V), where Tobit type I stands for the first model described above. Schnedler (2005) provides a general formula to obtain consistent likelihood estimators for these and other variations of the Tobit model.[7]

Type I[edit]

The Tobit model is a special case of a censored regression model, because the latent variable cannot always be observed while the independent variable is observable. A common variation of the Tobit model is censoring at a value different from zero:

Another example is censoring of values above .

Yet another model results when is censored from above and below at the same time.

The rest of the models will be presented as being bounded from below at 0, though this can be generalized as done for Type I.

Type II[edit]

Type II Tobit models introduce a second latent variable.

Heckman (1987) falls into the Type II Tobit. In Type I Tobit, the latent variable absorb both the process of participation and 'outcome' of interest. Type II Tobit allows the process of participation/selection and the process of 'outcome' to be independent, conditional on x.

Type III[edit]

Type III introduces a second observed dependent variable.

The Heckman model falls into this type.

Type IV[edit]

Type IV introduces a third observed dependent variable and a third latent variable.

Type V[edit]

Similar to Type II, in Type V only the sign of is observed.

The likelihood function[edit]

Below are the likelihood and log likelihood functions for a type I Tobit. This is a Tobit that is censored from below at when the latent variable . In writing out the likelihood function, we first define an indicator function where:

Next, let be the standard normal cumulative distribution function and to be the standard normal probability density function. For a data set with N observations the likelihood function for a type I Tobit is

and the log likelihood is given by

Note that this is different from the likelihood function of the truncated regression model.[8]

Non-Parametric Version[edit]

If the underlying latent variable is not normally distributed, one must use quantiles instead of moments to analyze the observable variable . Powell's CLAD estimator offers a possible way to achieve this.[9]

Dynamic Unobserved Effects Tobit Model[edit]

In a panel data Tobit model,[10][11] if the outcome partially depends on the previous outcome history this Tobit model is called "dynamic". For instance, taking a person who finds a job with a high salary this year, it will be easier for her to find a job with a high salary next year because the fact that she has a high-wage job this year will be a very positive signal for the potential employers. The essence of this type of dynamic effect is the state dependence of the outcome. The "unobservable effects" here refers to the factor which partially determines the outcome of individual but cannot be observed in the data. For instance, the ability of a person is very important in job-hunting, but it is not observable for researchers. A typical Dynamic Unobserved Effects Tobit Model can be represented as:

In this specific model, is the dynamic effect part and is the unobserved effect part whose distribution is determined by the initial outcome of individual i and some exogenous features of individual i.

Based on this setup, the likelihood function conditional on can be given as:

For the initial values ,there are two different ways to treat them in the construction of the likelihood function: treating them as constant, or imposing a distribution on them and calculate out the unconditional likelihood function. But whichever way is chosen to treat the initial values in the likelihood function, we cannot get rid of the integration inside the likelihood function when estimating the model by Maximum Likelihood Estimation (MLE). Expectation Maximum (EM) algorithm is usually a good solution for this computation issue.[12] Based on the consistent point estimates from MLE, Average Partial Effect (APE)[13] can be calculated correspondingly.[14]

Applications[edit]

Tobit models have, for example, been applied to estimate factors that impact grant receipt, including financial transfers distributed to sub-national governments who may apply for these grants. In these cases, grant recipients cannot receive negative amounts, and the data is thus left-censored. For instance, Dahlberg and Johansson (2002)[15] analyse a sample of 115 municipalities (42 of which received a grant). Dubois and Fattore (2011)[16] use a Tobit model to investigate the role of various factors in European Union fund receipt by applying Polish sub-national governments. The data may however be left-censored at a point higher than zero, with the risk of mis-specification. Both studies apply Probit and other models to check for robustness.

See also[edit]

References[edit]

  1. ^ Tobin, James (1958). "Estimation of relationships for limited dependent variables". Econometrica. 26 (1): 24–36. doi:10.2307/1907382. JSTOR 1907382. 
  2. ^ International Encyclopedia of the Social Sciences (2008)
  3. ^ Park, B.U., L. Simar, and V. Zelenyuk (2008). "Local likelihood estimation of truncated regression and its partial derivatives: Theory and application," Journal of Econometrics 146(1), pages 185-198.
  4. ^ The ET Interview: Professor James Tobin
  5. ^ Amemiya, Takeshi (1973). "Regression analysis when the dependent variable is truncated normal". Econometrica. 41 (6): 997–1016. doi:10.2307/1914031. JSTOR 1914031. 
  6. ^ McDonald, John F.; Moffit, Robert A. (1980). "The Uses of Tobit Analysis". The Review of Economics and Statistics. The MIT Press. 62 (2): 318–321. doi:10.2307/1924766. JSTOR 1924766. 
  7. ^ Schnedler, Wendelin (2005). "Likelihood estimation for censored random vectors". Econometric Reviews. 24 (2): 195–217. doi:10.1081/ETC-200067925. 
  8. ^ Park, B.U., L. Simar, and V. Zelenyuk (2008). "Local likelihood estimation of truncated regression and its partial derivatives: Theory and application," Journal of Econometrics 146(1), pages 185-198.
  9. ^ Powell, James L (1 July 1984). "Least absolute deviations estimation for the censored regression model". Journal of Econometrics. 25 (3): 303–325. doi:10.1016/0304-4076(84)90004-6. 
  10. ^ Greene, W. H. (2003), Econometric Analysis , Prentice Hall , Upper Saddle River, NJ.
  11. ^ The model framework comes from Wooldridge, J. (2002): Econometric Analysis of Cross Section and Panel Data, MIT Press, Cambridge, Mass, pp 542. But the author revises the model more general here.
  12. ^ For more details, refer to: Cappé, O., Moulines, E. and Ryden, T. (2005): “Inference in Hidden Markov Models”, Springer-Verlag New York, Chapter 2
  13. ^ Wooldridge, J. (2002): Econometric Analysis of Cross Section and Panel Data, MIT Press, Cambridge, Mass, pp 22.
  14. ^ For more details, refer to: Takeshi Amemiya (1984),”Tobit models: A survey”, Journal of Econometrics, 24, pp 3-6
  15. ^ Dahlberg, Matz; Johansson, Eva (2002-03-01). "On the Vote-Purchasing Behavior of Incumbent Governments". American Political Science Review. null (01): 27–40. doi:10.1017/S0003055402004215. ISSN 1537-5943. 
  16. ^ Dubois, Hans F. W.; Fattore, Giovanni (2011-07-01). "Public Fund Assignment through Project Evaluation". Regional & Federal Studies. 21 (3): 355–374. doi:10.1080/13597566.2011.578827. ISSN 1359-7566. 

Further reading[edit]