1,720,984 research outputs found
Maximum likelihood estimation of spatially and serially correlated panels with random effects
No full textAn estimation framework and a user-friendly software implementation are described for maximum likelihood estimation of panel data models with random effects, a spatially lagged dependent variable and spatially and serially correlated errors. This specification extends static panel data models in the direction of serial error correlation, allowing richer modelling possibilities and more thorough diagnostic assessments. The estimation routines extend the functionalities of the splm package for spatial panel econometrics in the open source R system for statistical computing
Specifying spatial effects in panel data: Locally robust vs. conditional tests
Full text linkWe address the issue of specifying a spatial lag vs. spatial error process in spatial panel models. The popular locally robust Lagrange multiplier (RLM) tests for spatial lag vs. error are compared to optimal alternatives based on maximum likelihood estimation: Wald and likelihood ratio (LR) tests requiring estimation of the full encompassing model, and conditional Lagrange multiplier (CLM) tests drawing on the reduced specification. Monte Carlo simulations are performed in a typical spatial panel context. Individual effects are successfully eliminated through the forward orthogonal deviations transformation, making the RLM suitable for panel data. Nevertheless, the statistical properties of Wald and LR are superior to those of the RLM. The CLM also dominates the RLM, as long as the sample is at least of moderate size. The RLM are computationally very convenient, but ML-based tests are feasible in most usage cases on mainstream hardware
The generalized spatial random effects model in R
Full text linkWe describe a user-friendly, production quality R implementation of the maximum likelihood estimator of the generalized spatial random effects (GSRE) model of Baltagi, Egger and Pfaffermayr within the well known ’splm’ package for spatial panel econometrics. We extend the maximum likelihood estimator for the GSRE to including a spatial lag of the dependent variable (SAR), and we discuss the theoretical and computational approach. This is the first implementation of the SAR+GSRE, and the second of the original GSRE. Until recently only estimators restricting the spatial structure of individual effects in an arbitrary way have been available and widely employed in applied practice. We present results from the SAR+GSRE and the restricted estimators side by side, drawing on some well-known examples from the spatial econometrics literature. The potential biases from imposing inappropriate restrictions to the spatial error process and/or from omitting the SAR term are illustrated by simulation
Narrow Replication of "A Spatio-Temporal Model of House Prices in the Usa" Using R
No full textI narrowly replicate Holly et al.'s (Journal of Econometrics 2010; 158(1): 160–173) analysis of the housing market in the USA, using the open source R software instead of the original ad hoc GAUSS routines. Their main findings are confirmed and most results are matched exactly. Attention is given to providing a self-contained and fully reproducible analysis, exclusively using user-level features available in the public domain
Life insurance in Italy: A sub-regional panel data analysis
No full textLavoro presentato ad "International Thematic workshop on Geographic Information Analysis for economic and spatial development and planning. Case studies, methods and models” (Geog An Mod 2010 GO Local), 3-4 Maggio 2010, Gorizia, Italia
Insurance consumption in Italy: a sub-regional panel data analysis
No full textLavoro presentato a SEA 2007, July 11-14 (The first World Conference of the Spatial Econometrics Association) - Cambridge (UK
splm : Spatial Panel Data Models in R
Full text linksplm is an R package for the estimation and testing of various spatial panel data specifications. We consider the implementation of both maximum likelihood and generalized moments estimators in the context of xed as well as random effects spatial panel data models. This paper is a general description of splm and all functionalities are illustrated using a well-known example taken from Munnell(1990)with productivity data on 48 US states observed over 17 years. We perform comparisons with other available software; and, when this is not possible, Monte Carlo results support our original implementation
Are funding of pensions and economic growth directly linked? New empirical results for some OECD countries
No full textWe empirically test on a panel of OECD countries the hypothesis of a direct and positive link between funding of pensions and economic growth, which is based on the idea that richer pension systems can accelerate the development of the financial system and thus promote a more efficient capital allocation. We follow Davis and Hu (2008) [Davis and Hu (2008), Does funding of pensions stimulate economic growth?, Journal of Pension Economic and Finance, Cambridge University Press, vol. 7 (02), 221-249] in estimating a modified Cobb-Douglas production function, where pension fund assets are treated as a shift factor, but we criticize their results from an econometric point of view, since both the Dynamic OLS and Mean Group (MG) estimators are inadequate in case of cross-sectionally correlated residuals. Indeed, we find a highly significant level of correlation in the MG residuals across countriesthat we attribute to common global shocks driving per capita outputs. Therefore we adopt a more general approach suitable to the presence of a multifactor error structure. Our results exclude the existence of a long run cointegration relationship between autonomous (or total) pension fund assets and per capita output for our panel of OECD countries, unless, in contrast to the conclusion of the cross-sectional dependence test, we ignore it and assume independence of residuals
Non-life insurance consumption in Italy: a sub-regional panel data analysis
No full textInsurance development, Regional panel, Spatial dependence, C31, C33, G22,
Insurance in Italy: a spatial perspective
No full textThe authors analyze the demand for life and non-life insurance across 103 Italian provinces in 1994-2004, assessing the determinants of insurance consumption in the light of the empirical literature and the distinctive features of the Italian market, which we thoroughly describe. The authors discuss common issues in the empirical literature on insurance development, presenting the sub-regional perspective as a partial solution; at the same time, they elaborate on the peculiar issues arising from the use of sub-regional data: spatial heterogeneity and spatial correlation. They describe the evolution of provincial heterogeneity and of the spatial features of the data over the observation period. In order to control for both unobserved heterogeneity and spatial correlation, they specify a spatial panel model with random provincial effects and macroregional fixed effects, which is estimated by maximum likelihood. The chapter carefully assesses the properties of model residuals, concluding that the specification allows for reliable inference on the drivers of insurance consumption. It concludes describing the empirical findings and giving some suggestions for future research
- …
