Panel Data III. Stefan Dahlberg

Size: px
Start display at page:

Download "Panel Data III. Stefan Dahlberg"

Transcription

1 Panel Data III Stefan Dahlberg

2 Overview of Part #3 VII. Spatial Panel Data & Unit Heterogeneity VIII. Panel Corrected Standard Errors (PCSE) IX. Fixed and Random Effects X. Hausman Test XI. The Hybrid Model XII. Working with TSCS data in Stata (cont.) c. Commands tailored for panel data

3 Potential Problems with Panel Data 1. Autocorrelation in the time dimension within units Expected effects: biased standard errors 2. Spatial autocorrelation (correlated errors between units that are geographically proximate) Expected effects: ineffient β:s, biased standard errors 3. Heteroscedasticity (i.e. unequal variance between units) Expected effects: ineffient β:s, biased standard errors 4. Contemporaneous correlation of errors across units (something affects all units at the same time) Expected effects: ineffient β:s, biased standard errors 5. Structural issues (likely the case that each disparate unit will have a unique constant as each unit likely has a distinct history) Expected effects: inefficient β:s, biased standard errors

4 VII. Spatial Panel Data and Unit Heterogeneity

5 Additional estimation problems Spatial autocorrelation (r[e i,t,e j,t ] 0): X i,t Y i,t Y j,t e.g., diffusion process usually a problem when geography is at hand.

6 Spatial autocorrelation (r[e i,t,e j,t ] 0):

7 Panel heteroskedasticity Panel heteroscedasticity or cluster heterogeneity (V[e i ] σ for all i) (e.g., quality of data varies with GDP/cap) implies that cluster means of the DV varies across clusters due to unmeasured cluster-level factors. Unobserved heterogeneity should always be addressed in clustered data (such as panel or multilevel data). For some models, one can include observed variables that will explain parts of the between cluster differences (instead of country dummies you control for GDP, political system, population etc).

8 Panel heteroskedasticity (Figure from Bartels 2009)

9 Correlated Residuals Larger Residuals as Y increases - Heteroscedasticity of Errors

10 Additional estimation problems How detect heteroscedasticity? Use same tools as for simple OLS (Breusch-Pagan or White s test). Breusch-Pagan test: Regress full model (no xtset is needed) hettest indep. vars. gives a chi2-squared test statistic (H0=homoscedasticity in the error terms - if Prob>chi2=<.05 H0 is rejected heteroscedasticity is at hand)

11 Three strategies for Paneldata 1. OLS Completely Pooled * - TSCS (Data :N>T) use autoregressive model with PCSE or go for: 2. Fixed Effects (FE) Unit-Specific Eff. No pooling (Data: N<=T) 3. Random Effects (RE) Partial pooling (Data: N<=T) The last two approaches account for unobserved heterogeneity, although in different ways, while the complete pooling strategy ignores the unobserved heterogeneity (when PCSE:s are not used) *all units are characterized by the same regression equation at all points in time.

12 Pooled OLS

13 OLS with PCSE:s Panel-corrected standard errors (PCSE s - Beck & Katz -APSR 1995)), by assuming Correlations between residuals vary across pairs of units, but remain constant over time within each pair and are only contemporaneous (r[e i,t,e j,t ]= r[e i,t,e j,t ] 0, but r[e i,t,e j,t ]=0) The variance of the residuals vary by country but remain constant over time* (V[e i,t ] V[e j,t ], but V[e i,t ]=V[e j,t ]) Stata estimates these correlations and variances from T the data and corrects for them (xtpcse) ^ ei, te j, t t 1 i, j * correlation in residuals over time is corrected by LDV T

14 Panel Corrected Standard Errors Beck and Katz propose an estimator that pools information across clusters to estimate the error variances. The Beck & Katz panel corrected standard error is calculated in the following way. Organize the residuals from the fitted model (OLS) according to cluster, so that the residuals from the clusters are ˆe 1 ˆe 2,..., ˆe N. These are vectors with T elements each, and they can be grouped together and used for weighting residuals across units. as a T N matrix (the ˆe i are columns): In Stata: xtpcse instead of xtreg

15 What is Panel Corrected Standard Errors? Aimed at estimating correlations across units. Such as contemporaneous correlation and additional problem of shared error effects across units panel heteroscedasticity. Why not always use PCSE:s? When homoskedasticity assumption holds and errors are normally distributed the t- statistics have an exact t-distribution even when sample sizes are small and with small samples the t-statistics for corrected or robust std. err. are not very close to the t- distrubution. (Wooldrige 2003:261)

16 Panel Corrected Standard Errors The pcse standard error estimate is robust not only to unit heteroskedacity, but it also robust against possible contemporaneous correlation across the units that is common in TSCS data. Note that using panel-corrected standard errors (PCSE:s Beck & Katz 1995) is a complete pooling approach that does not account for unobserved heterogeneity. The PCSE:s do, of course, make corrections for the standard errors, but the OLS coefficients are completely pooled estimates. The major payoff of this approach is its simplicity while a disadvantage of complete pooling is that it ignores unobserved Heterogeneity which can induce omitted variable bias (Skrondal and Rabe-Hesketh 2004)

17 VIII. Fixed- and Random Effects

18 Fixed Effects (FE) Fixed-effects (FE) also known as Unit Specific Effects - are used when only interested in analyzing the impact of variables that vary over time. Since fixed effects estimators depends on deviations from their group means, they are sometimes referred to as within-groups estimators (Davidson and MacKinnon, 1993). Equation for the FE model is: Yit = αi + βxit + uit Where: αi (i=1.n) is the unknown intercept for each unit (n unit-specific intercepts). Yit is the dep. var. (DV) where i = unit and t = time. Xit represents one indep. var. (IV) β1 is the coefficient for that IV, uit is the error term

19 Fixed Effects (FE) (Figure from Bartels 2009)

20 Fixed Effects (FE) FE explore the relationship between indep.vars and dep.vars within a unit (country, person, organization, etc.). Each unit has its own characteristics that may or may not influence the indep. vars (gender could for example influence opinion on certain issue or the electoral system of a specific country may affect party competition or a company's business practices may influence its stock price). When using FE we assume that something within the individual/unit may impact or bias the indep. or dep. vars and we need to control for this. This is the logic behind the assumption of the correlation between a units error term and indep. variables. FE remove the effect of time-invariant characteristics from the indep. vars.

21 Fixed Effects (FE) Fixed effects models are not without their drawbacks. The fixed effects models may have too many cross-sectional units of observations requiring too many dummy variables for model specification. Too many variables decreases the degrees of freedom for adequately statistical tests. A model with too many variables may be plagued with multicollinearity, which increases the standard errors. If the model contain variables that do not vary within the groups - parameter estimation may be inefficient. Autocorrelation over time (serial-correlation) is not solved by FE. Use LDV if T>15 or use Praise-regression

22 Fixed Effects (FE) FE models cannot be used to investigate time-invariant causes of the dependent variables, this since time-invariant characteristics of the units are perfectly collinear with the dummies. The FE approach thus eliminates the ability to test between-cluster hypotheses Fixed-effects will not work well with data for which within-cluster variation is minimal or for slow changing variables over time. Since all between-cluster variation in the data is absorbed by the cluster-specific dummies, the effects of independent variables are solely within-cluster effects, which has implications for how one interprets coefficients. For TSCS data, such effects are interpreted as: for a given country, as X varies across time by one unit, Y increases or decreases by ᵝ units (Bartels 2009).

23 For TSCS data, a now standard modeling practice is to use an FE model with panel-corrected standard errors and a lagged dependent variable to account for dynamics (Beck and Katz 1996; Beck 2001; Wilson and Butler 2007), though there is not an ironclad consensus about this strategy among practitioners (Bartels 2009) However fixed effects with a lagged DV does only perform well when t>15 (approximately)

24 AN EXAMPLE OF WHEN IT DOES MATTER

25 There can be little doubt when viewed as a single-level model, that the relation is a negative one, albeit one with quite a lot of scatter around the line. Turning now to a random intercept multilevel model (Goldstein 2003), we can recognise that the observations belong to 10 groups.

26 With single level regression, it is assumed that the observations are independently and identically distributed, and this gives an overall negative relationship. There is no recognition that within each group the underlying relationship is positive. In contrast the multilevel model allows the intercept of each group to take on a different value from an overall distribution. As the following table shows, the multilevel model is a much better fit to these data, with a considerably smaller deviance. The multilevel model gives a substantially better interpretation of the data. There is no reason why such relationships should not be found in reality. The effect of taking account of groups in the multilevel model is marked here because the group-specific intercept is negatively related to the mean of X for each group. This behaviour can be elucidated by including the group mean of X in the multilevel model alongside the deviations of X from that mean (Paccagnella, 2006).

27 It is now clear that the between-group relation between Y and Xbar is markedly negative, while the within-group relation with X-Xbar is positive. The true relation between Y and X is only revealed when the within- and between-group relations are considered jointly in a multilevel model. 6

28 FE in Stata With OLS and dummies xi: reg Y Xk i.id With xtreg and fe-option xtset id time xtreg Y Xk, fe Note: sometimes you might need to convert id to numeric, type: encode id, gen(id2). Use id2 instead of id in the xtset command

29 Testing for Heteroskedasticity A test for heteroskedasticiy available for fixed-effect models is: In Stata: xttest3 The null-hypothesis is homoskedasticity (constant variance). If we reject the null-hypothesis heteroskedasticity is a problem. Solution: Use the option robust to obtain heteroskedasticityrobust standard errors (Huber/White or sandwich estimators).

30 Random Effects (RE) The rationale behind random effects model is that, unlike the fixed effects model, the variation across entities is assumed to be random and uncorrelated with the predictor or independent variables included in the model (Torres-Reyna 200x)). Yit = α + βxit + uit + εit Between cluster error Within cluster error If you have reason to believe that differences across entities have some influence on your dependent variable then you should use random effects. An advantage of random effects is that you can include time invariant variables (i.e. gender). In the fixed effects model these variables are absorbed by the intercept variation.

31 Assumptions of RE The two components of the composite error, Ui and εit are independent, i.e. E(Uε)=0 2. The variances of both Ui (σ2 u) and εit (σ2 ε) are constant for all X (no heteroskedasticity) 3. The idiosyncratic residuals εit at one point in time are not related to their value at another point in time (no autocorrelation in εit). These three are relatively unproblematic but: 4. Both Ui and εit are unrelated to the Xik, i.e. E(XU)=E(Xε)=0 In order to use a RE model to identify and estimate a β with two separate error terms, they need to be treated as unrelated to the observed independent variables. This is the common OLS assumption for ε, but is now extend to U as well. Otherwise we can t estimate the separate effects of X and the composite error.

32 Since a level-1 variable varies both within and between clusters, many argue that this an unrealistic assumption to satisfy, since unobserved heterogeneity will almost always be correlated with the independent variables (Bartels 2009). Assumptions of RE A random effects, or random intercept, approach treats u0j as distributed normally with mean zero and an estimable variance. This approach decomposes the total error into a level-1 component (eij) and a level-2 component (u0j). The RE model is a partial pooling approach, with the effects of X1ij and X2ij a weighted average of the within and between-cluster variation in the data (Gelman and Hill 2007). A major complaint lodged against the RE model relates to the restrictive assumption that level-1 independent variables be uncorrelated with the random effects term on lev. 2: Cov(Xij, u0j)=0.

33 If assumption (4) is true, then RE is definitely the best estimator available. How does the RE estimator work, what its advantages

34 How to estimate RE How do we estimate a model such as the equation below? Yit= α + β1x1it +β2x2it +β3x3it+...βkxikt +Ui +ε it

35 RE in Stata With xtreg and re-option xtset id time xtreg Y Xk, re xtmixed Y Xk id

36 Mean of country level intercepts xtreg fh_polity2, re Std. Dev of country level intercepts Std. Dev. At the time level

37 But what is it really? How to estimate RE Several ways, but the simplest is to estimate by Generalized Least Squares (GLS), which involves weighting the equation by a factor that will transform the problematic error term so that OLS can be used on the weighted or transformed model. This is what used to be the most common correction for heteroskedasticity: weight the data by the inverse of X or the square of X (because the unequal variance was an increasing or decreasing function of X), and this weighted equation would then have an error variance that satisfied OLS assumptions.so WLS (Weighted Least Squares) is a type of GLS estimation. Another example of GLS is in time-series analysis, where one might weight the data by ρ, the autocorrelation parameter for the εit, and then use OLS on the weighted data to estimate structural effects when the error term is autocorrelated. Generally: GLS proceeds by weighting the data by the inverse of the error variance-covariance matrix to ensure that the weighted equation has a normal structure with common variance on the diagonals and zero covariances on the off-diagonals. Then OLS is used on the weighted equation.

38 ICC Intraclass Correlation The proportion of the total variability in the outcome that is attributed to the higher levels. Calculating the ICC show how much of the variability in the dependent variable that is attributed to the Higher Level Clusters (e.g. how much is attributed to variation between countries/regions/schools and how much is between individuals within these units) High ICC high variability at the highest level. Low ICC low variability at the highest level.

39 ICC Intraclass Correlation Amount of Variance in DV explained by the unit-specific (level 2) variables ICC Variance in Random Intercepts Between Group Variance Intercept [Subject = Variance Variable Name] We need to model unit-effects if this number is higher than 0.05 Total Variance = Variance in Random Intercepts + Residual Variance Between Group Variance + Within Group Variance Intercept [Subject = Variance Variable Name] + Residual

40 ICC Intraclass Correlation If the intra-class correlation approaches 1, there is no variance to explain at the individual level all units are similar results are driven by btw cluster effects If ICC is 0 = no clustering structure use simple OLS ICC Total Variance = Variance in Random Intercepts + Residual Variance Between Group Variance + Within Group Variance Intercept [Subject = Variance Variable Name] + Residual

41 Needed: estimates of the two variance terms σ2 u + σ2 e. If we could obtain those estimates, we can weight or transform equation in the following way and then use OLS to estimate the effects: Yit θ Yi = (α θα ) +β1(x1it θ X1i ) +β2 (X2it θ X2i ) +βk (Xikt θ Xik ) + (Uit θui + ε it θεi ) If the observed Y and X in the model are transformed/weighted by equations θ ( Theta ) then the resulting error term in will be OLS-ready. If we knew σ2 u + σ2 e. we could simply use them directly but we don t know the population values of these two error variances. We need to estimate them from our data, which is why this application is called Feasible Generalized Least Squares (FGLS)

42 How do we get estimates of σ2u + σ2e? From the Within (FE) Regression we get an estimate of σ2 e. Why? FE eliminates Ui altogether, and the error term is pure ε From the Between Regression (the means of Yi over time against the means of all the Xi over time), we get an error term whose variance is: or the variance of U plus the time-averaged variance of ε. With this information we can calculate an estimate of each error component, calculate θ or THETA, transform the equation by this estimate and re-estimate the model with OLS Computational Formula for θ:

43 So what does the RE actually do? We can examine θ more closely to get a better idea of what RE is doing. As θ (theta) gets closer to 1, it means that more and more of the composite error variance is made up of Ui unit-level or between variance. So what happens then? Then the weighted RE equation (13) reduces to the FE equation! because if all of the error variation is from U, let s difference out U completely as the FE model does. As theta gets closer to 0, it means that more of more of the composite error variance is made up of random idiosyncratic variance ε, with no unit variance at all. So the RE equation reduces to POOLED OLS in this instance!! This is also as it should be because we only should take into account unit effects when they exist!

44 So what does the RE actually do? So we can look at the RE estimator as a weighted average of FE and pooled OLS, with the weight (θ) depending on how much of the estimated composite error variance is from the units. This is a middle ground, then, between the full unitlevel differencing model of FE and the assumption of no unit effects in pooled OLS. If there is a lot of unit-level variation, then RE is closer to FE. If there is not so much unit-level variation, then RE is closer to pooled OLS. This seems reasonable, IF the RE assumption of zero correlation between X and U is tenable a big if!!

45 IX. Hausman test

46 Fixed or Random Effects? To decide between fixed or random effects you can run a Hausman (1978) test where the null hypothesis is that the preferred model is random effects vs. the alternative the fixed effects (see Green, 2008). Hausman is thus used to detect violations of the random-effects modeling assumption that the explanatory variables are orthogonal to the unit effects. It basically tests whether the unique errors (ui) are correlated with the regressors, the null hypothesis is they are not. A significant test result is taken as evidence of a correlation between x and αj, implying that the random-effects model should be rejected in favor of the fixed-effects model.

47 Fixed or Random Effects? How to conduct a Hausman test to detect violations of the randomeffects modeling assumption: Run a fixed effects model and save the estimates, then run a random model and save the estimates and perform the test, such as: xtreg y x1, fe estimates store fixed xtreg y x1, re estimates store random hausman fixed random

48 Command reg xtreg areg Syntax Entity fixed effects reg y x1 x2 x3 x4 x5 i.country xtreg y x1 x2 x3 x4 x5, fe areg y x1 x2 x3 x4 x5, absorb(country) reg xtreg areg Entity and time fixed effects reg y x1 x2 x3 x4 x5 i.country i.year xtreg y x1 x2 x3 x4 x5 i.year, fe areg y x1 x2 x3 x4 x5 i.year, absorb(country) xtreg xtreg Random Effects xtreg y x1 x2 x3 x4 x5, re xtreg y x1 x2 x3 x4 x5, re robust

49 The Hybrid Model What to do if we want time-invariant covariates but the RE-assumption doesn t hold? To keep the FE set up, while trying to say something about the effects of time-invariant Xs; and keep the RE set up, while at the same time allowing possible correlation between the X and the Ui.

50 Rabe-Hesketh&Skrondal/Bell&Jones:Hybrid Model The idea is that the possible covariation of time-varying Xs and the Ui is what messes up RE. But this possible covariation is the result of model misspecification - something in the Ui term is related to the X that we need to account for, and RE cannot account for it due to its assumption that E(XU)=0. But we can bring the covariation between X and Ui into the model indirectly, by including the mean of X as an additional independent variable in Whatever covariation between X and U that may exist is now accounted for; if units that are generally high (low) on X also have high (low) U terms, then the mean of X in the model will pick this up. The effect of regular X can now be estimated, controlling for this possible confounding problem.

51 Rabe-Hesketh&Skrondal/Bell&Jones:Hybrid Model in Stata use "C:\Steve\exercise2.dta", clear gen mad_gdppcl=ln(mad_gdppc) egen gdpmean=mean(mad_gdppcl), by (ccode) gen gdpmeandev=mad_gdppcl-gdpmean xtreg fh_ipolity2 gdpmeandev gdpmean al_ethnic al_religion, re xtreg fh_ipolity2 gdpmeandev gdpmean al_ethnic al_religion, re vce(cluster ccode)

52 Potential Problems and Solutions with Panel Data Analysis 1. Autocorrelation in the time dimension within units Lagged DV, Praise-Winsten or Cochrane Orcutt 2. Spatial autocorrelation (correlated errors between units that are geographically proximate) PCSE:s 3. Heteroscedasticity (i.e. unequal variance between units) FE or RE (PCSE) 4. Contemporaneous correlation of errors across units (something affects all units at the same time) PCSE 5. Structural issues (likely the case that each disparate unit will have a unique constant as each unit likely has a distinct history) FE or RE

53 Panel data. Heterogeneity in levels and effects So far we talked about the inherent problems with the data. Other (and equally serious) Sources of heterogeneity that inflate the previously noted problems Intercept Heterogeneity Assuming a common intercept is as problematic as assuming constant error variance. Is it reasonable to assume that Welfare spending levels to be the same in Sweden and in Spain, absent all other factors that are relevant? Slope Heterogeneity Assuming common slopes can lead to erroneous inferences between dependent and independent variables. The effect of X on Y might be accelerated due to context specific effects that are not being accounted for in the model. Simultaneous intercept and slope heterogeneity

54 1. OLS Interce pt Slope Error term within clusters Variability in Intercepts between groups 2. Random Intercept Variability in Slopes within clusters 3. Random Coefficient 4. Random Coefficient And Random Intercept

55 Panel data. Heterogeneity in levels and effects Heterogeneity (example from Wilson and Butler, 2007)

56 EXERCISE 5 1. Open the abridged Russian panel data set in WIDE format: exercise1 2. Run the following regression: supdem i,t = α+γ supdem i,t-1 + β lifesat i,t-1 + e i,t 3. Open the same data set in LONG format: exercise3 4. Run the same regression and compare results 5. Experiment with the lag length of lifesat and various controls (gender, age, education) 6. Run the regression from your final model but with PCSE:s, FE and RE (3 models) and compare results

57 SOLUTION use "exercise1", clear regr supdem2 supdem1 v39y1 regr supdem3 supdem2 v39y1 use "exercise3", clear xtset v1 wave regr supdem l.supdem l.v39y regr supdem l.supdem v39y l.v39y regr supdem l.supdem v39y l.v39y v35x v42x v293x set matsize 1500 xtpcse supdem l.supdem v39y l.v39y v35x v42x v293x, p xtreg supdem l.supdem v39y l.v39y v35x v42x v293x, fe xtreg supdem l.supdem v39y l.v39y v35x v42x v293x, re

58 EXERCISE #6 1. Open the abridged QoG panel data set from file exercise2 2. Run a autoregressiv equation with fh_ipolity2 i,t as the dependent variable and choose two independent variables 3. Run the same regression as in 2 but with pcse:s 4. Run the same regression as in 2 but with fixed effects using xtreg and fe-option 5. Run the same regression as in 2 but with fixed effects using country dummies and compare results from 3 and 4. Run the same regression as in 2 but with fixed effects and PCSE:s 6. Conduct a Hausman test 7. Run the same regression but now using random effects.

59 Solution use "exercise2", clear xtset ccode year xtreg fh_ipolity2 l.fh_ipolity2 l.mad_gdppc l.mad_pop set matsize 1500 xtpcse fh_ipolity2 l.fh_ipolity2 l.mad_gdppc l.mad_pop, p xtreg fh_ipolity2 l.fh_ipolity2 l.mad_gdppc l.mad_pop, fe xi: reg fh_ipolity2 l.fh_ipolity2 l.mad_gdppc l.mad_pop i.ccode xtpcse fh_ipolity2 l.fh_ipolity2 l.mad_gdppc l.mad_pop i.ccode, p xtreg fh_ipolity2 l.fh_ipolity2 l.mad_gdppc l.mad_pop, fe estimates store fixed xtreg fh_ipolity2 l.fh_ipolity2 l.mad_gdppc l.mad_pop, re estimates store random hausman fixed random xtreg fh_ipolity2 l.fh_ipolity2 l.mad_gdppc l.mad_pop, re

60 Cross-sectional dependence According to Baltagi, cross-sectional dependence is a problem in macro panels with long time series (over years). This is not much of a problem in micro panels (few years and large number of cases). The null hypothesis in the B-P/LM test of independence is that residuals across units are not correlated. In Stata xttest2 (run it after xtreg, fe): xtreg y x1, fe xttest2

Capital humain, développement et migrations: approche macroéconomique (Empirical Analysis - Static Part)

Capital humain, développement et migrations: approche macroéconomique (Empirical Analysis - Static Part) Séminaire d Analyse Economique III (LECON2486) Capital humain, développement et migrations: approche macroéconomique (Empirical Analysis - Static Part) Frédéric Docquier & Sara Salomone IRES UClouvain

More information

Panel Data. March 2, () Applied Economoetrics: Topic 6 March 2, / 43

Panel Data. March 2, () Applied Economoetrics: Topic 6 March 2, / 43 Panel Data March 2, 212 () Applied Economoetrics: Topic March 2, 212 1 / 43 Overview Many economic applications involve panel data. Panel data has both cross-sectional and time series aspects. Regression

More information

Time-Series Cross-Section Analysis

Time-Series Cross-Section Analysis Time-Series Cross-Section Analysis Models for Long Panels Jamie Monogan University of Georgia February 17, 2016 Jamie Monogan (UGA) Time-Series Cross-Section Analysis February 17, 2016 1 / 20 Objectives

More information

Econometrics of Panel Data

Econometrics of Panel Data Econometrics of Panel Data Jakub Mućk Meeting # 4 Jakub Mućk Econometrics of Panel Data Meeting # 4 1 / 30 Outline 1 Two-way Error Component Model Fixed effects model Random effects model 2 Non-spherical

More information

Recent Advances in the Field of Trade Theory and Policy Analysis Using Micro-Level Data

Recent Advances in the Field of Trade Theory and Policy Analysis Using Micro-Level Data Recent Advances in the Field of Trade Theory and Policy Analysis Using Micro-Level Data July 2012 Bangkok, Thailand Cosimo Beverelli (World Trade Organization) 1 Content a) Classical regression model b)

More information

ECON 4551 Econometrics II Memorial University of Newfoundland. Panel Data Models. Adapted from Vera Tabakova s notes

ECON 4551 Econometrics II Memorial University of Newfoundland. Panel Data Models. Adapted from Vera Tabakova s notes ECON 4551 Econometrics II Memorial University of Newfoundland Panel Data Models Adapted from Vera Tabakova s notes 15.1 Grunfeld s Investment Data 15.2 Sets of Regression Equations 15.3 Seemingly Unrelated

More information

Longitudinal Data Analysis Using Stata Paul D. Allison, Ph.D. Upcoming Seminar: May 18-19, 2017, Chicago, Illinois

Longitudinal Data Analysis Using Stata Paul D. Allison, Ph.D. Upcoming Seminar: May 18-19, 2017, Chicago, Illinois Longitudinal Data Analysis Using Stata Paul D. Allison, Ph.D. Upcoming Seminar: May 18-19, 217, Chicago, Illinois Outline 1. Opportunities and challenges of panel data. a. Data requirements b. Control

More information

Repeated observations on the same cross-section of individual units. Important advantages relative to pure cross-section data

Repeated observations on the same cross-section of individual units. Important advantages relative to pure cross-section data Panel data Repeated observations on the same cross-section of individual units. Important advantages relative to pure cross-section data - possible to control for some unobserved heterogeneity - possible

More information

Econ 1123: Section 5. Review. Internal Validity. Panel Data. Clustered SE. STATA help for Problem Set 5. Econ 1123: Section 5.

Econ 1123: Section 5. Review. Internal Validity. Panel Data. Clustered SE. STATA help for Problem Set 5. Econ 1123: Section 5. Outline 1 Elena Llaudet 2 3 4 October 6, 2010 5 based on Common Mistakes on P. Set 4 lnftmpop = -.72-2.84 higdppc -.25 lackpf +.65 higdppc * lackpf 2 lnftmpop = β 0 + β 1 higdppc + β 2 lackpf + β 3 lackpf

More information

Basic Regressions and Panel Data in Stata

Basic Regressions and Panel Data in Stata Developing Trade Consultants Policy Research Capacity Building Basic Regressions and Panel Data in Stata Ben Shepherd Principal, Developing Trade Consultants 1 Basic regressions } Stata s regress command

More information

Topic 10: Panel Data Analysis

Topic 10: Panel Data Analysis Topic 10: Panel Data Analysis Advanced Econometrics (I) Dong Chen School of Economics, Peking University 1 Introduction Panel data combine the features of cross section data time series. Usually a panel

More information

INTRODUCTION TO BASIC LINEAR REGRESSION MODEL

INTRODUCTION TO BASIC LINEAR REGRESSION MODEL INTRODUCTION TO BASIC LINEAR REGRESSION MODEL 13 September 2011 Yogyakarta, Indonesia Cosimo Beverelli (World Trade Organization) 1 LINEAR REGRESSION MODEL In general, regression models estimate the effect

More information

Please discuss each of the 3 problems on a separate sheet of paper, not just on a separate page!

Please discuss each of the 3 problems on a separate sheet of paper, not just on a separate page! Econometrics - Exam May 11, 2011 1 Exam Please discuss each of the 3 problems on a separate sheet of paper, not just on a separate page! Problem 1: (15 points) A researcher has data for the year 2000 from

More information

Cross Sectional Time Series: The Normal Model and Panel Corrected Standard Errors

Cross Sectional Time Series: The Normal Model and Panel Corrected Standard Errors Cross Sectional Time Series: The Normal Model and Panel Corrected Standard Errors Paul Johnson 5th April 2004 The Beck & Katz (APSR 1995) is extremely widely cited and in case you deal

More information

Topic 7: Heteroskedasticity

Topic 7: Heteroskedasticity Topic 7: Heteroskedasticity Advanced Econometrics (I Dong Chen School of Economics, Peking University Introduction If the disturbance variance is not constant across observations, the regression is heteroskedastic

More information

Applied Microeconometrics (L5): Panel Data-Basics

Applied Microeconometrics (L5): Panel Data-Basics Applied Microeconometrics (L5): Panel Data-Basics Nicholas Giannakopoulos University of Patras Department of Economics ngias@upatras.gr November 10, 2015 Nicholas Giannakopoulos (UPatras) MSc Applied Economics

More information

PS 271B: Quantitative Methods II Lecture Notes

PS 271B: Quantitative Methods II Lecture Notes PS 271B: Quantitative Methods II Lecture Notes (Part 6: Panel/Longitudinal Data; Multilevel/Mixed Effects models) Langche Zeng zeng@ucsd.edu Panel/Longitudinal Data; Multilevel Modeling; Mixed effects

More information

EC327: Advanced Econometrics, Spring 2007

EC327: Advanced Econometrics, Spring 2007 EC327: Advanced Econometrics, Spring 2007 Wooldridge, Introductory Econometrics (3rd ed, 2006) Chapter 14: Advanced panel data methods Fixed effects estimators We discussed the first difference (FD) model

More information

Econometrics. Week 4. Fall Institute of Economic Studies Faculty of Social Sciences Charles University in Prague

Econometrics. Week 4. Fall Institute of Economic Studies Faculty of Social Sciences Charles University in Prague Econometrics Week 4 Institute of Economic Studies Faculty of Social Sciences Charles University in Prague Fall 2012 1 / 23 Recommended Reading For the today Serial correlation and heteroskedasticity in

More information

Econometrics. 9) Heteroscedasticity and autocorrelation

Econometrics. 9) Heteroscedasticity and autocorrelation 30C00200 Econometrics 9) Heteroscedasticity and autocorrelation Timo Kuosmanen Professor, Ph.D. http://nomepre.net/index.php/timokuosmanen Today s topics Heteroscedasticity Possible causes Testing for

More information

Chapter 15 Panel Data Models. Pooling Time-Series and Cross-Section Data

Chapter 15 Panel Data Models. Pooling Time-Series and Cross-Section Data Chapter 5 Panel Data Models Pooling Time-Series and Cross-Section Data Sets of Regression Equations The topic can be introduced wh an example. A data set has 0 years of time series data (from 935 to 954)

More information

Advanced Econometrics

Advanced Econometrics Based on the textbook by Verbeek: A Guide to Modern Econometrics Robert M. Kunst robert.kunst@univie.ac.at University of Vienna and Institute for Advanced Studies Vienna May 16, 2013 Outline Univariate

More information

Iris Wang.

Iris Wang. Chapter 10: Multicollinearity Iris Wang iris.wang@kau.se Econometric problems Multicollinearity What does it mean? A high degree of correlation amongst the explanatory variables What are its consequences?

More information

Fortin Econ Econometric Review 1. 1 Panel Data Methods Fixed Effects Dummy Variables Regression... 7

Fortin Econ Econometric Review 1. 1 Panel Data Methods Fixed Effects Dummy Variables Regression... 7 Fortin Econ 495 - Econometric Review 1 Contents 1 Panel Data Methods 2 1.1 Fixed Effects......................... 2 1.1.1 Dummy Variables Regression............ 7 1.1.2 First Differencing Methods.............

More information

Applied Economics. Panel Data. Department of Economics Universidad Carlos III de Madrid

Applied Economics. Panel Data. Department of Economics Universidad Carlos III de Madrid Applied Economics Panel Data Department of Economics Universidad Carlos III de Madrid See also Wooldridge (chapter 13), and Stock and Watson (chapter 10) 1 / 38 Panel Data vs Repeated Cross-sections In

More information

FinQuiz Notes

FinQuiz Notes Reading 10 Multiple Regression and Issues in Regression Analysis 2. MULTIPLE LINEAR REGRESSION Multiple linear regression is a method used to model the linear relationship between a dependent variable

More information

10 Panel Data. Andrius Buteikis,

10 Panel Data. Andrius Buteikis, 10 Panel Data Andrius Buteikis, andrius.buteikis@mif.vu.lt http://web.vu.lt/mif/a.buteikis/ Introduction Panel data combines cross-sectional and time series data: the same individuals (persons, firms,

More information

Lab 11 - Heteroskedasticity

Lab 11 - Heteroskedasticity Lab 11 - Heteroskedasticity Spring 2017 Contents 1 Introduction 2 2 Heteroskedasticity 2 3 Addressing heteroskedasticity in Stata 3 4 Testing for heteroskedasticity 4 5 A simple example 5 1 1 Introduction

More information

Econometrics of Panel Data

Econometrics of Panel Data Econometrics of Panel Data Jakub Mućk Meeting # 6 Jakub Mućk Econometrics of Panel Data Meeting # 6 1 / 36 Outline 1 The First-Difference (FD) estimator 2 Dynamic panel data models 3 The Anderson and Hsiao

More information

Lecture 9: Panel Data Model (Chapter 14, Wooldridge Textbook)

Lecture 9: Panel Data Model (Chapter 14, Wooldridge Textbook) Lecture 9: Panel Data Model (Chapter 14, Wooldridge Textbook) 1 2 Panel Data Panel data is obtained by observing the same person, firm, county, etc over several periods. Unlike the pooled cross sections,

More information

Analysis of Panel Data: Fixed Effects, Random Effects, and Hybrid Models for Multi-Wave Panel Data

Analysis of Panel Data: Fixed Effects, Random Effects, and Hybrid Models for Multi-Wave Panel Data Analysis of Panel Data: Fixed Effects, Random Effects, and Hybrid Models for Multi-Wave Panel Data Session 2: 15 June 2015 Steven Finkel, PhD Daniel Wallace Professor of Political Science University of

More information

multilevel modeling: concepts, applications and interpretations

multilevel modeling: concepts, applications and interpretations multilevel modeling: concepts, applications and interpretations lynne c. messer 27 october 2010 warning social and reproductive / perinatal epidemiologist concepts why context matters multilevel models

More information

Empirical Application of Panel Data Regression

Empirical Application of Panel Data Regression Empirical Application of Panel Data Regression 1. We use Fatality data, and we are interested in whether rising beer tax rate can help lower traffic death. So the dependent variable is traffic death, while

More information

Advanced Quantitative Methods: panel data

Advanced Quantitative Methods: panel data Advanced Quantitative Methods: Panel data University College Dublin 17 April 2012 1 2 3 4 5 Outline 1 2 3 4 5 Panel data Country Year GDP Population Ireland 1990 80 3.4 Ireland 1991 84 3.5 Ireland 1992

More information

Econometrics of Panel Data

Econometrics of Panel Data Econometrics of Panel Data Jakub Mućk Meeting # 2 Jakub Mućk Econometrics of Panel Data Meeting # 2 1 / 26 Outline 1 Fixed effects model The Least Squares Dummy Variable Estimator The Fixed Effect (Within

More information

Jeffrey M. Wooldridge Michigan State University

Jeffrey M. Wooldridge Michigan State University Fractional Response Models with Endogenous Explanatory Variables and Heterogeneity Jeffrey M. Wooldridge Michigan State University 1. Introduction 2. Fractional Probit with Heteroskedasticity 3. Fractional

More information

Ninth ARTNeT Capacity Building Workshop for Trade Research "Trade Flows and Trade Policy Analysis"

Ninth ARTNeT Capacity Building Workshop for Trade Research Trade Flows and Trade Policy Analysis Ninth ARTNeT Capacity Building Workshop for Trade Research "Trade Flows and Trade Policy Analysis" June 2013 Bangkok, Thailand Cosimo Beverelli and Rainer Lanz (World Trade Organization) 1 Selected econometric

More information

An overview of applied econometrics

An overview of applied econometrics An overview of applied econometrics Jo Thori Lind September 4, 2011 1 Introduction This note is intended as a brief overview of what is necessary to read and understand journal articles with empirical

More information

Introduction to Panel Data Analysis

Introduction to Panel Data Analysis Introduction to Panel Data Analysis Youngki Shin Department of Economics Email: yshin29@uwo.ca Statistics and Data Series at Western November 21, 2012 1 / 40 Motivation More observations mean more information.

More information

Econometrics. Week 8. Fall Institute of Economic Studies Faculty of Social Sciences Charles University in Prague

Econometrics. Week 8. Fall Institute of Economic Studies Faculty of Social Sciences Charles University in Prague Econometrics Week 8 Institute of Economic Studies Faculty of Social Sciences Charles University in Prague Fall 2012 1 / 25 Recommended Reading For the today Instrumental Variables Estimation and Two Stage

More information

Lecture 5: Omitted Variables, Dummy Variables and Multicollinearity

Lecture 5: Omitted Variables, Dummy Variables and Multicollinearity Lecture 5: Omitted Variables, Dummy Variables and Multicollinearity R.G. Pierse 1 Omitted Variables Suppose that the true model is Y i β 1 + β X i + β 3 X 3i + u i, i 1,, n (1.1) where β 3 0 but that the

More information

Econometrics of Panel Data

Econometrics of Panel Data Econometrics of Panel Data Jakub Mućk Meeting # 4 Jakub Mućk Econometrics of Panel Data Meeting # 4 1 / 26 Outline 1 Two-way Error Component Model Fixed effects model Random effects model 2 Hausman-Taylor

More information

Panel Data Models. Chapter 5. Financial Econometrics. Michael Hauser WS17/18 1 / 63

Panel Data Models. Chapter 5. Financial Econometrics. Michael Hauser WS17/18 1 / 63 1 / 63 Panel Data Models Chapter 5 Financial Econometrics Michael Hauser WS17/18 2 / 63 Content Data structures: Times series, cross sectional, panel data, pooled data Static linear panel data models:

More information

Econometrics of Panel Data

Econometrics of Panel Data Econometrics of Panel Data Jakub Mućk Meeting # 1 Jakub Mućk Econometrics of Panel Data Meeting # 1 1 / 31 Outline 1 Course outline 2 Panel data Advantages of Panel Data Limitations of Panel Data 3 Pooled

More information

CRE METHODS FOR UNBALANCED PANELS Correlated Random Effects Panel Data Models IZA Summer School in Labor Economics May 13-19, 2013 Jeffrey M.

CRE METHODS FOR UNBALANCED PANELS Correlated Random Effects Panel Data Models IZA Summer School in Labor Economics May 13-19, 2013 Jeffrey M. CRE METHODS FOR UNBALANCED PANELS Correlated Random Effects Panel Data Models IZA Summer School in Labor Economics May 13-19, 2013 Jeffrey M. Wooldridge Michigan State University 1. Introduction 2. Linear

More information

Economics 308: Econometrics Professor Moody

Economics 308: Econometrics Professor Moody Economics 308: Econometrics Professor Moody References on reserve: Text Moody, Basic Econometrics with Stata (BES) Pindyck and Rubinfeld, Econometric Models and Economic Forecasts (PR) Wooldridge, Jeffrey

More information

Week 11 Heteroskedasticity and Autocorrelation

Week 11 Heteroskedasticity and Autocorrelation Week 11 Heteroskedasticity and Autocorrelation İnsan TUNALI Econ 511 Econometrics I Koç University 27 November 2018 Lecture outline 1. OLS and assumptions on V(ε) 2. Violations of V(ε) σ 2 I: 1. Heteroskedasticity

More information

Quantitative Methods Final Exam (2017/1)

Quantitative Methods Final Exam (2017/1) Quantitative Methods Final Exam (2017/1) 1. Please write down your name and student ID number. 2. Calculator is allowed during the exam, but DO NOT use a smartphone. 3. List your answers (together with

More information

Problem Set 10: Panel Data

Problem Set 10: Panel Data Problem Set 10: Panel Data 1. Read in the data set, e11panel1.dta from the course website. This contains data on a sample or 1252 men and women who were asked about their hourly wage in two years, 2005

More information

Applied Statistics and Econometrics

Applied Statistics and Econometrics Applied Statistics and Econometrics Lecture 6 Saul Lach September 2017 Saul Lach () Applied Statistics and Econometrics September 2017 1 / 53 Outline of Lecture 6 1 Omitted variable bias (SW 6.1) 2 Multiple

More information

Short T Panels - Review

Short T Panels - Review Short T Panels - Review We have looked at methods for estimating parameters on time-varying explanatory variables consistently in panels with many cross-section observation units but a small number of

More information

point estimates, standard errors, testing, and inference for nonlinear combinations

point estimates, standard errors, testing, and inference for nonlinear combinations Title xtreg postestimation Postestimation tools for xtreg Description The following postestimation commands are of special interest after xtreg: command description xttest0 Breusch and Pagan LM test for

More information

FNCE 926 Empirical Methods in CF

FNCE 926 Empirical Methods in CF FNCE 926 Empirical Methods in CF Lecture 11 Standard Errors & Misc. Professor Todd Gormley Announcements Exercise #4 is due Final exam will be in-class on April 26 q After today, only two more classes

More information

LECTURE 10. Introduction to Econometrics. Multicollinearity & Heteroskedasticity

LECTURE 10. Introduction to Econometrics. Multicollinearity & Heteroskedasticity LECTURE 10 Introduction to Econometrics Multicollinearity & Heteroskedasticity November 22, 2016 1 / 23 ON PREVIOUS LECTURES We discussed the specification of a regression equation Specification consists

More information

Introduction to Econometrics. Heteroskedasticity

Introduction to Econometrics. Heteroskedasticity Introduction to Econometrics Introduction Heteroskedasticity When the variance of the errors changes across segments of the population, where the segments are determined by different values for the explanatory

More information

Pooling Space and Time

Pooling Space and Time Pooling Space and Time Jamie Monogan University of Georgia March 21, 2012 Jamie Monogan (UGA) Pooling Space and Time March 21, 2012 1 / 47 Objectives By the end of this meeting participants should be able

More information

Statistics, inference and ordinary least squares. Frank Venmans

Statistics, inference and ordinary least squares. Frank Venmans Statistics, inference and ordinary least squares Frank Venmans Statistics Conditional probability Consider 2 events: A: die shows 1,3 or 5 => P(A)=3/6 B: die shows 3 or 6 =>P(B)=2/6 A B : A and B occur:

More information

Econometrics for PhDs

Econometrics for PhDs Econometrics for PhDs Amine Ouazad April 2012, Final Assessment - Answer Key 1 Questions with a require some Stata in the answer. Other questions do not. 1 Ordinary Least Squares: Equality of Estimates

More information

Semester 2, 2015/2016

Semester 2, 2015/2016 ECN 3202 APPLIED ECONOMETRICS 5. HETEROSKEDASTICITY Mr. Sydney Armstrong Lecturer 1 The University of Guyana 1 Semester 2, 2015/2016 WHAT IS HETEROSKEDASTICITY? The multiple linear regression model can

More information

Gov 2000: 9. Regression with Two Independent Variables

Gov 2000: 9. Regression with Two Independent Variables Gov 2000: 9. Regression with Two Independent Variables Matthew Blackwell Fall 2016 1 / 62 1. Why Add Variables to a Regression? 2. Adding a Binary Covariate 3. Adding a Continuous Covariate 4. OLS Mechanics

More information

Econometric Analysis of Cross Section and Panel Data

Econometric Analysis of Cross Section and Panel Data Econometric Analysis of Cross Section and Panel Data Jeffrey M. Wooldridge / The MIT Press Cambridge, Massachusetts London, England Contents Preface Acknowledgments xvii xxiii I INTRODUCTION AND BACKGROUND

More information

Wooldridge, Introductory Econometrics, 4th ed. Chapter 15: Instrumental variables and two stage least squares

Wooldridge, Introductory Econometrics, 4th ed. Chapter 15: Instrumental variables and two stage least squares Wooldridge, Introductory Econometrics, 4th ed. Chapter 15: Instrumental variables and two stage least squares Many economic models involve endogeneity: that is, a theoretical relationship does not fit

More information

The Simple Linear Regression Model

The Simple Linear Regression Model The Simple Linear Regression Model Lesson 3 Ryan Safner 1 1 Department of Economics Hood College ECON 480 - Econometrics Fall 2017 Ryan Safner (Hood College) ECON 480 - Lesson 3 Fall 2017 1 / 77 Bivariate

More information

Wooldridge, Introductory Econometrics, 2d ed. Chapter 8: Heteroskedasticity In laying out the standard regression model, we made the assumption of

Wooldridge, Introductory Econometrics, 2d ed. Chapter 8: Heteroskedasticity In laying out the standard regression model, we made the assumption of Wooldridge, Introductory Econometrics, d ed. Chapter 8: Heteroskedasticity In laying out the standard regression model, we made the assumption of homoskedasticity of the regression error term: that its

More information

ECONOMETRICS HONOR S EXAM REVIEW SESSION

ECONOMETRICS HONOR S EXAM REVIEW SESSION ECONOMETRICS HONOR S EXAM REVIEW SESSION Eunice Han ehan@fas.harvard.edu March 26 th, 2013 Harvard University Information 2 Exam: April 3 rd 3-6pm @ Emerson 105 Bring a calculator and extra pens. Notes

More information

1 The Multiple Regression Model: Freeing Up the Classical Assumptions

1 The Multiple Regression Model: Freeing Up the Classical Assumptions 1 The Multiple Regression Model: Freeing Up the Classical Assumptions Some or all of classical assumptions were crucial for many of the derivations of the previous chapters. Derivation of the OLS estimator

More information

Panel Data: Fixed and Random Effects

Panel Data: Fixed and Random Effects Short Guides to Microeconometrics Fall 2016 Kurt Schmidheiny Unversität Basel Panel Data: Fixed and Random Effects 1 Introduction In panel data, individuals (persons, firms, cities, ) are observed at several

More information

Econometrics Multiple Regression Analysis: Heteroskedasticity

Econometrics Multiple Regression Analysis: Heteroskedasticity Econometrics Multiple Regression Analysis: João Valle e Azevedo Faculdade de Economia Universidade Nova de Lisboa Spring Semester João Valle e Azevedo (FEUNL) Econometrics Lisbon, April 2011 1 / 19 Properties

More information

ECON2228 Notes 10. Christopher F Baum. Boston College Economics. cfb (BC Econ) ECON2228 Notes / 48

ECON2228 Notes 10. Christopher F Baum. Boston College Economics. cfb (BC Econ) ECON2228 Notes / 48 ECON2228 Notes 10 Christopher F Baum Boston College Economics 2014 2015 cfb (BC Econ) ECON2228 Notes 10 2014 2015 1 / 48 Serial correlation and heteroskedasticity in time series regressions Chapter 12:

More information

Graduate Econometrics Lecture 4: Heteroskedasticity

Graduate Econometrics Lecture 4: Heteroskedasticity Graduate Econometrics Lecture 4: Heteroskedasticity Department of Economics University of Gothenburg November 30, 2014 1/43 and Autocorrelation Consequences for OLS Estimator Begin from the linear model

More information

Econometrics - 30C00200

Econometrics - 30C00200 Econometrics - 30C00200 Lecture 11: Heteroskedasticity Antti Saastamoinen VATT Institute for Economic Research Fall 2015 30C00200 Lecture 11: Heteroskedasticity 12.10.2015 Aalto University School of Business

More information

Internal vs. external validity. External validity. This section is based on Stock and Watson s Chapter 9.

Internal vs. external validity. External validity. This section is based on Stock and Watson s Chapter 9. Section 7 Model Assessment This section is based on Stock and Watson s Chapter 9. Internal vs. external validity Internal validity refers to whether the analysis is valid for the population and sample

More information

Applied Econometrics. Lecture 3: Introduction to Linear Panel Data Models

Applied Econometrics. Lecture 3: Introduction to Linear Panel Data Models Applied Econometrics Lecture 3: Introduction to Linear Panel Data Models Måns Söderbom 4 September 2009 Department of Economics, Universy of Gothenburg. Email: mans.soderbom@economics.gu.se. Web: www.economics.gu.se/soderbom,

More information

statistical sense, from the distributions of the xs. The model may now be generalized to the case of k regressors:

statistical sense, from the distributions of the xs. The model may now be generalized to the case of k regressors: Wooldridge, Introductory Econometrics, d ed. Chapter 3: Multiple regression analysis: Estimation In multiple regression analysis, we extend the simple (two-variable) regression model to consider the possibility

More information

Econometrics Summary Algebraic and Statistical Preliminaries

Econometrics Summary Algebraic and Statistical Preliminaries Econometrics Summary Algebraic and Statistical Preliminaries Elasticity: The point elasticity of Y with respect to L is given by α = ( Y/ L)/(Y/L). The arc elasticity is given by ( Y/ L)/(Y/L), when L

More information

Multiple Regression Analysis: Heteroskedasticity

Multiple Regression Analysis: Heteroskedasticity Multiple Regression Analysis: Heteroskedasticity y = β 0 + β 1 x 1 + β x +... β k x k + u Read chapter 8. EE45 -Chaiyuth Punyasavatsut 1 topics 8.1 Heteroskedasticity and OLS 8. Robust estimation 8.3 Testing

More information

Session 3-4: Estimating the gravity models

Session 3-4: Estimating the gravity models ARTNeT- KRI Capacity Building Workshop on Trade Policy Analysis: Evidence-based Policy Making and Gravity Modelling for Trade Analysis 18-20 August 2015, Kuala Lumpur Session 3-4: Estimating the gravity

More information

Heteroskedasticity in Panel Data

Heteroskedasticity in Panel Data Essex Summer School in Social Science Data Analysis Panel Data Analysis for Comparative Research Heteroskedasticity in Panel Data Christopher Adolph Department of Political Science and Center for Statistics

More information

Heteroskedasticity in Panel Data

Heteroskedasticity in Panel Data Essex Summer School in Social Science Data Analysis Panel Data Analysis for Comparative Research Heteroskedasticity in Panel Data Christopher Adolph Department of Political Science and Center for Statistics

More information

Making sense of Econometrics: Basics

Making sense of Econometrics: Basics Making sense of Econometrics: Basics Lecture 4: Qualitative influences and Heteroskedasticity Egypt Scholars Economic Society November 1, 2014 Assignment & feedback enter classroom at http://b.socrative.com/login/student/

More information

LECTURE 10: MORE ON RANDOM PROCESSES

LECTURE 10: MORE ON RANDOM PROCESSES LECTURE 10: MORE ON RANDOM PROCESSES AND SERIAL CORRELATION 2 Classification of random processes (cont d) stationary vs. non-stationary processes stationary = distribution does not change over time more

More information

THE MULTIVARIATE LINEAR REGRESSION MODEL

THE MULTIVARIATE LINEAR REGRESSION MODEL THE MULTIVARIATE LINEAR REGRESSION MODEL Why multiple regression analysis? Model with more than 1 independent variable: y 0 1x1 2x2 u It allows : -Controlling for other factors, and get a ceteris paribus

More information

Econometrics. Week 6. Fall Institute of Economic Studies Faculty of Social Sciences Charles University in Prague

Econometrics. Week 6. Fall Institute of Economic Studies Faculty of Social Sciences Charles University in Prague Econometrics Week 6 Institute of Economic Studies Faculty of Social Sciences Charles University in Prague Fall 2012 1 / 21 Recommended Reading For the today Advanced Panel Data Methods. Chapter 14 (pp.

More information

Longitudinal Data Analysis Using SAS Paul D. Allison, Ph.D. Upcoming Seminar: October 13-14, 2017, Boston, Massachusetts

Longitudinal Data Analysis Using SAS Paul D. Allison, Ph.D. Upcoming Seminar: October 13-14, 2017, Boston, Massachusetts Longitudinal Data Analysis Using SAS Paul D. Allison, Ph.D. Upcoming Seminar: October 13-14, 217, Boston, Massachusetts Outline 1. Opportunities and challenges of panel data. a. Data requirements b. Control

More information

PANEL DATA RANDOM AND FIXED EFFECTS MODEL. Professor Menelaos Karanasos. December Panel Data (Institute) PANEL DATA December / 1

PANEL DATA RANDOM AND FIXED EFFECTS MODEL. Professor Menelaos Karanasos. December Panel Data (Institute) PANEL DATA December / 1 PANEL DATA RANDOM AND FIXED EFFECTS MODEL Professor Menelaos Karanasos December 2011 PANEL DATA Notation y it is the value of the dependent variable for cross-section unit i at time t where i = 1,...,

More information

The gravity models for trade research

The gravity models for trade research The gravity models for trade research ARTNeT-CDRI Capacity Building Workshop Gravity Modelling 20-22 January 2015 Phnom Penh, Cambodia Dr. Witada Anukoonwattaka Trade and Investment Division, ESCAP anukoonwattaka@un.org

More information

Panel data methods for policy analysis

Panel data methods for policy analysis IAPRI Quantitative Analysis Capacity Building Series Panel data methods for policy analysis Part I: Linear panel data models Outline 1. Independently pooled cross sectional data vs. panel/longitudinal

More information

Econometrics I KS. Module 2: Multivariate Linear Regression. Alexander Ahammer. This version: April 16, 2018

Econometrics I KS. Module 2: Multivariate Linear Regression. Alexander Ahammer. This version: April 16, 2018 Econometrics I KS Module 2: Multivariate Linear Regression Alexander Ahammer Department of Economics Johannes Kepler University of Linz This version: April 16, 2018 Alexander Ahammer (JKU) Module 2: Multivariate

More information

STOCKHOLM UNIVERSITY Department of Economics Course name: Empirical Methods Course code: EC40 Examiner: Lena Nekby Number of credits: 7,5 credits Date of exam: Saturday, May 9, 008 Examination time: 3

More information

Motivation for multiple regression

Motivation for multiple regression Motivation for multiple regression 1. Simple regression puts all factors other than X in u, and treats them as unobserved. Effectively the simple regression does not account for other factors. 2. The slope

More information

ECON2228 Notes 10. Christopher F Baum. Boston College Economics. cfb (BC Econ) ECON2228 Notes / 54

ECON2228 Notes 10. Christopher F Baum. Boston College Economics. cfb (BC Econ) ECON2228 Notes / 54 ECON2228 Notes 10 Christopher F Baum Boston College Economics 2014 2015 cfb (BC Econ) ECON2228 Notes 10 2014 2015 1 / 54 erial correlation and heteroskedasticity in time series regressions Chapter 12:

More information

Review of Panel Data Model Types Next Steps. Panel GLMs. Department of Political Science and Government Aarhus University.

Review of Panel Data Model Types Next Steps. Panel GLMs. Department of Political Science and Government Aarhus University. Panel GLMs Department of Political Science and Government Aarhus University May 12, 2015 1 Review of Panel Data 2 Model Types 3 Review and Looking Forward 1 Review of Panel Data 2 Model Types 3 Review

More information

1 The basics of panel data

1 The basics of panel data Introductory Applied Econometrics EEP/IAS 118 Spring 2015 Related materials: Steven Buck Notes to accompany fixed effects material 4-16-14 ˆ Wooldridge 5e, Ch. 1.3: The Structure of Economic Data ˆ Wooldridge

More information

Dynamic Panel Data Ch 1. Reminder on Linear Non Dynamic Models

Dynamic Panel Data Ch 1. Reminder on Linear Non Dynamic Models Dynamic Panel Data Ch 1. Reminder on Linear Non Dynamic Models Pr. Philippe Polomé, Université Lumière Lyon M EcoFi 016 017 Overview of Ch. 1 Data Panel Data Models Within Estimator First-Differences Estimator

More information

A Course in Applied Econometrics Lecture 7: Cluster Sampling. Jeff Wooldridge IRP Lectures, UW Madison, August 2008

A Course in Applied Econometrics Lecture 7: Cluster Sampling. Jeff Wooldridge IRP Lectures, UW Madison, August 2008 A Course in Applied Econometrics Lecture 7: Cluster Sampling Jeff Wooldridge IRP Lectures, UW Madison, August 2008 1. The Linear Model with Cluster Effects 2. Estimation with a Small Number of roups and

More information

Applied Quantitative Methods II

Applied Quantitative Methods II Applied Quantitative Methods II Lecture 10: Panel Data Klára Kaĺıšková Klára Kaĺıšková AQM II - Lecture 10 VŠE, SS 2016/17 1 / 38 Outline 1 Introduction 2 Pooled OLS 3 First differences 4 Fixed effects

More information

Econometrics of Panel Data

Econometrics of Panel Data Econometrics of Panel Data Jakub Mućk Meeting # 3 Jakub Mućk Econometrics of Panel Data Meeting # 3 1 / 21 Outline 1 Fixed or Random Hausman Test 2 Between Estimator 3 Coefficient of determination (R 2

More information

Panel Data: Very Brief Overview Richard Williams, University of Notre Dame, https://www3.nd.edu/~rwilliam/ Last revised April 6, 2015

Panel Data: Very Brief Overview Richard Williams, University of Notre Dame, https://www3.nd.edu/~rwilliam/ Last revised April 6, 2015 Panel Data: Very Brief Overview Richard Williams, University of Notre Dame, https://www3.nd.edu/~rwilliam/ Last revised April 6, 2015 These notes borrow very heavily, often verbatim, from Paul Allison

More information

Econ 836 Final Exam. 2 w N 2 u N 2. 2 v N

Econ 836 Final Exam. 2 w N 2 u N 2. 2 v N 1) [4 points] Let Econ 836 Final Exam Y Xβ+ ε, X w+ u, w N w~ N(, σi ), u N u~ N(, σi ), ε N ε~ Nu ( γσ, I ), where X is a just one column. Let denote the OLS estimator, and define residuals e as e Y X.

More information

1 Motivation for Instrumental Variable (IV) Regression

1 Motivation for Instrumental Variable (IV) Regression ECON 370: IV & 2SLS 1 Instrumental Variables Estimation and Two Stage Least Squares Econometric Methods, ECON 370 Let s get back to the thiking in terms of cross sectional (or pooled cross sectional) data

More information