Non-Tariff Barriers to Agricultural Trade between Turkey and the EU
(Sprache: Englisch)
This thesis reviews the border effect approach as an application of gravity models of trade and different methods of including multilateral resistance terms (MRTs) in it. Some focus is laid on the endogeneity problem of the approach. In an empirical...
Voraussichtlich lieferbar in 3 Tag(en)
versandkostenfrei
Buch (Kartoniert)
Fr. 56.00
inkl. MwSt.
- Kreditkarte, Paypal, Rechnungskauf
- 30 Tage Widerrufsrecht
Produktdetails
Produktinformationen zu „Non-Tariff Barriers to Agricultural Trade between Turkey and the EU “
Klappentext zu „Non-Tariff Barriers to Agricultural Trade between Turkey and the EU “
This thesis reviews the border effect approach as an application of gravity models of trade and different methods of including multilateral resistance terms (MRTs) in it. Some focus is laid on the endogeneity problem of the approach. In an empirical application of the approach, agricultural trade between Turkey and the EU is analysed; the effect of data pooling and aggregation is studied; the conversion of estimated border effects into ad-valorem tariff equivalents (AVEs) reveals the crucial importance of a reliable measure of the elasticity of substitution when trying to separate the effects of NTBs of the total effect of a border.
Lese-Probe zu „Non-Tariff Barriers to Agricultural Trade between Turkey and the EU “
Chapter 2.3.4; Maximum Likelihood Estimation and Poisson Regressions:Maximum likelihood estimation aims to find estimates of the coefficients that make the observed data most likely to occur; i.e. it is assumed that the observed data is only a sample of all possible observations, and then those values of the coefficients that would make the observed data most likely to occur in a random sample are chosen. For this purpose, a function that explains the likelihood of the observed data depending on the estimated coefficients is created and maximised iteratively. In practice, it is generally more feasible to take the logarithm of that likelihood function, multiply it by minus one and then minimise it. (Purcell 2007)
Alternatively, a simplified form of the log-likelihood function may be used. In that case, the estimation is called pseudo- or quasi-maximum likelihood. (Gong and Samaniego 1981)
The PPML estimator employed by Santos Silva and Tenreyro (2006) applies pseudo-maximum likelihood estimator to Poisson distributed numbers.
A Poisson distribution predicts the likelihood of an integer, nonnegative count variable, such as the number of pieces of mail a person receives each day, taking a certain value, given that the mean of the assumed values is known and that assumed values do not depend on previously assumed values. (Bruce Brooks 2001)
A Poisson regression, then, models the likelihood of a dependent count variable assuming a certain value as a function of explanatory variables, and maximises that likelihood. It has the property of yielding consistent and asymptotically normal estimators for the explanatory variables coefficients even if the dependent variable does not follow a Poisson distribution. Thus, when a Poisson regression is run on a dependent variable that may not be Poisson distributed, it can be called a quasi-maximum likelihood estimation (QMLE). (Wooldridge 2009, p. 597-598)
An alternative term for quasi-maximum likelihood estimation in this
... mehr
context is pseudo-maximum likelihood estimation. (Lindsay 1988)
2.3.5 Application of the PPML Estimator in the Gravity Model:
Santos Silva and Tenreyro (2010) apply the PPML estimator in their gravity model: Based on the assumption that the conditional variance is proportional to the conditional mean, i.e. E[y_i x] V[y_i x], can be estimated solving a set of first-order conditions: _(i=1)^n [y_i-exp (x_i )] x_i =0
The PML estimator gives the same weight to all observations, because observations additional information on curvature of the conditional mean is countered by their larger variance. This is a difference to the NLS estimator, which can be interpreted as a PML estimator assuming that V[y_i x]=const. . While there is no further information available on heteroskedasticity, giving the same weight to all observations appears appropriate.
The authors find the estimator defined by equation (35) to be numerically equal to the PPML estimator. Further, they infer from equation (35) that the only prerequisite for the PPML estimator to be consistent is the correct specification of the conditional mean, i.e. E[y_i x]=exp (x_i ). Therefore, even if the data is not Poisson distributed and yi is not an integer, the estimator defined by equation (35) is still numerically equal to the PPML estimator.
This proves to be particularly handy since the PPML estimator can easily be calculated in standard econometric programmes such as STATA, the one used for this thesis.
In general, if it was known that V[yi x] is a function of higher powers of E[yi x], more efficient estimates could be achieved by weighing even less the large observations. In the case of trade data, however, the data of larger countries can be assumed to be of higher quality than that of smaller countries, which is why it is probably better to not weigh small observations even more.
The PPML estimator, then, appears to be a good compromise, since it doesn t weigh the large observations with their high varianc
2.3.5 Application of the PPML Estimator in the Gravity Model:
Santos Silva and Tenreyro (2010) apply the PPML estimator in their gravity model: Based on the assumption that the conditional variance is proportional to the conditional mean, i.e. E[y_i x] V[y_i x], can be estimated solving a set of first-order conditions: _(i=1)^n [y_i-exp (x_i )] x_i =0
The PML estimator gives the same weight to all observations, because observations additional information on curvature of the conditional mean is countered by their larger variance. This is a difference to the NLS estimator, which can be interpreted as a PML estimator assuming that V[y_i x]=const. . While there is no further information available on heteroskedasticity, giving the same weight to all observations appears appropriate.
The authors find the estimator defined by equation (35) to be numerically equal to the PPML estimator. Further, they infer from equation (35) that the only prerequisite for the PPML estimator to be consistent is the correct specification of the conditional mean, i.e. E[y_i x]=exp (x_i ). Therefore, even if the data is not Poisson distributed and yi is not an integer, the estimator defined by equation (35) is still numerically equal to the PPML estimator.
This proves to be particularly handy since the PPML estimator can easily be calculated in standard econometric programmes such as STATA, the one used for this thesis.
In general, if it was known that V[yi x] is a function of higher powers of E[yi x], more efficient estimates could be achieved by weighing even less the large observations. In the case of trade data, however, the data of larger countries can be assumed to be of higher quality than that of smaller countries, which is why it is probably better to not weigh small observations even more.
The PPML estimator, then, appears to be a good compromise, since it doesn t weigh the large observations with their high varianc
... weniger
Bibliographische Angaben
- Autor: Claus Mayer
- 2014, Erstauflage, 56 Seiten, 5 Abbildungen, Masse: 15,5 x 22 cm, Kartoniert (TB), Englisch
- Verlag: Anchor Academic Publishing
- ISBN-10: 395489260X
- ISBN-13: 9783954892600
Sprache:
Englisch
Kommentar zu "Non-Tariff Barriers to Agricultural Trade between Turkey and the EU"
0 Gebrauchte Artikel zu „Non-Tariff Barriers to Agricultural Trade between Turkey and the EU“
Zustand | Preis | Porto | Zahlung | Verkäufer | Rating |
---|
Schreiben Sie einen Kommentar zu "Non-Tariff Barriers to Agricultural Trade between Turkey and the EU".
Kommentar verfassen