Should I Use Fixed or Random Effects?

Clark, Tom; Linzer, Drew A.

doi:10.1017/psrm.2014.32

Cited by 554 publications

(370 citation statements)

References 11 publications

Supporting

Mentioning

341

Contrasting

Unclassified

Order By: Relevance

“…However, econometricians now argue that the Hausman test does not help in deciding between fixed and random effects. Instead, more weight is given to the size of the data set, the extent of variability within units and the level of correlation between and within the covariates and units (Clark and Linzer 2012). The modelling of random effects usually follows strong assumptions, such as a normal distribution and independence of residuals of explanatory variables (without omitted variables), while fixed effects estimators do not rely on as strong assumptions as the random effects model and therefore are not likely to fail.…”

Section: Resultsmentioning

confidence: 99%

The politics of redistribution in local governments: the effect of gender representation on welfare spending in California counties

Park¹

2014

J. Pub. Pol.

View full text Add to dashboard Cite

: This research explores the impact of gender representation at the state and local levels on redistributive choices. This research also examines whether female officeholders moderate the impact of the local economy and institution on welfare spending. Hypotheses are tested across 58 counties in California over ten years, between 2001 and 2010. According to the fixed effect models, women in state legislature had a positive effect on local welfare spending, while women on county boards had no significant effect. However, a positive moderating effect of women on county boards during economic hardship was found. Three categories of control variables include institutional factors, such as the introduction of Proposition 1A and county home rule; political factors, such as the political preference of each county's residents and strength of non-profit organisations; and socio-economic factors, such as intergovernmental revenue, unemployment rate and demographics. Counties with more intergovernmental revenue and supporters of Democratic presidential candidates are likely to spend more on welfare services.

show abstract

Section: Resultsmentioning

confidence: 99%

The politics of redistribution in local governments: the effect of gender representation on welfare spending in California counties

Park¹

2014

J. Pub. Pol.

View full text Add to dashboard Cite

show abstract

“…The model selection relies on Monte Carlo studies, identifying it as the recommended option for small cross-sectional samples with a short time dimension (Petersen, 2009;Clark and Linzer, 2013), but our results are robust to estimating random effects models, and to not clustering the standard errors.…”

Section: Modelmentioning

confidence: 88%

The manipulation of basel risk-weights

Mariathasan

Merrouche

2014

Journal of Financial Intermediation

224

137

View full text Add to dashboard Cite

The Manipulation of Basel Risk-Weights* In this paper, we examine the relationship between banks' approval for the internal ratings-based (IRB) approaches of Basel II and the ratio of riskweighted over total assets. Analysing a panel of 115 banks from 21 OECD countries that were eventually approved for applying the IRB to their credit portfolio, we find that risk-weight density is lower once regulatory approval is granted. The effect persists when we control for different loan categories, and we provide evidence showing that it cannot be explained by flawed modelling, or improved risk-measurement alone. Consistent with theories of risk-weight manipulation, we find the decline in risk-weights to be particularly prevalent among weakly capitalised banks, when the legal framework for supervision is weak, and in countries where supervisors are overseeing many IRB banks. We conclude that part of the decline in reported riskiness under the IRB results from banks' strategic risk-modelling.JEL Classification: G20, G21 and G28

show abstract

“…Finally, as suggested by Clark, Liner (2015), different specifications of the model have been computed in order to confirm the stability of the results. As can be seen from Table A.1 (Appendix) the signs of the estimates for all parameters are the same independently of the type of model (fixed effects, random effects or pooling), although some differences can be observed regarding the statistical relevance of the estimates.…”

Section: Panel Data Model Without Spatial Effectsmentioning

confidence: 98%

“…Nevertheless, as discussed by Clark, Liner (2015), the Hausman test has important limitations for a final decision regarding the choice of a specific model, which should be grounded on theoretical assumptions about the observations. In our case -and considering the close link between the specific characteristics of the territories and tourism dynamics, it seems plausible to assume that individual regional features have specific impacts on tourism activities, also justifying the option for a fixed effects model.…”

Section: Panel Data Model Without Spatial Effectsmentioning

confidence: 99%

Territory and Sustainable Tourism Development: a Space-Time Analysis on European Regions

Romão

Guerreiro

Rodrigues

2017

REGION

View full text Add to dashboard Cite

Abstract. In the long run, tourism competitiveness depends on the sustainable use of territorial assets: the differentiation of destinations depends on the integration of cultural and natural resources into the tourism supply, but also on their preservation over time. Using advanced spatial econometric techniques this work analyses the relationships between regional tourism competitiveness, the dynamics of tourism demand and investment, as well as the existence of natural resources and cultural assets in the European context. Despite the close relationship between tourism activities and the characteristics of the territories, the application of methods of spatial analysis in tourism studies is still scarce and the results of this work clearly show their potential for this field of research. Among the main findings of this paper, it was observed that natural resources do not have the expected positive impacts on regional tourism competitiveness and that European regions with more abundant natural resources are often developing unsustainable forms of mass tourism with low value added and little benefits for the host communities. The existence of spatial correlation effects suggests that positive spillovers arising from tourism dynamics in neighbourhood regions prevail over potential negative effects related to the competition between destinations. Policy and managerial implications of these results are discussed and further research questions are proposed.

show abstract

Should I Use Fixed or Random Effects?

Cited by 554 publications

References 11 publications

The politics of redistribution in local governments: the effect of gender representation on welfare spending in California counties

The politics of redistribution in local governments: the effect of gender representation on welfare spending in California counties

The manipulation of basel risk-weights

Territory and Sustainable Tourism Development: a Space-Time Analysis on European Regions

Contact Info

Product

Resources

About