Vasilis Sarafidis' educational background is in the areas of economics and econometrics. He received his BA in Economics with Computing and Quantitative Methods at the University of Sussex, and his MPhil in Economics and Econometrics at the University of Cambridge. He completed his PhD in Econometrics in 2006 at the University of Cambridge. Prior to joining BI, he was Associate Professor at the Department of Econometrics and Business Statistics, Monash University, Melbourne.
Vasilis' main area of research lies in the econometric analysis of panel data. His focus lies on model specification and testing, dynamic panels, common factor structures, models for spatial interactions and networks, multi-dimensional panels, and random coefficient models; with applications to the analysis of crime, banking, water-usage demand, among others.
Before studying for his PhD, Vasilis worked as an economist for a private consultancy in the City of London. During that time he led projects involving econometric analysis and modelling for the electricity, water and telecommunications sectors. As a consultant he advised, among others, the Office for Water Services (U.K.), the National Telecom Agency in Denmark, the Jersey Competition Regulatory Authority and the Electricity Association (U.K.)
During his academic career, he has also acted as an advisor to the Productivity Commission (Australia), the Essential Services Commission in Victoria, the Commonwealth Environmental Office, the NSW Independent Pricing and Regulatory Tribunal, the Bureau of Airline Representatives of Australia, and the NSW Corrective Services.
Research areas Panel data analysis; model specification and testing; high-dimensional data; models for spatial interactions and networks.
Teaching areas GRA 6039 Econometrics with Programming
In this article, we introduce the xtgrangert command, which implements the panel Granger noncausality testing approach developed by Juodis, Karavias, and Sarafidis (2021, Empirical Economics 60: 93–112). This test offers superior size and power performance to existing tests, which stem from the use of a pooled estimator that has a faster NT−−−√ convergence rate. The test has several other useful properties: it can be used in multivariate systems; it has power against both homogeneous and heterogeneous alternatives; and it allows for cross-section dependence and cross-section heteroskedasticity.
This paper analyses the instrumental variables (IV) approach put forward by Norkute et al. (2021), in the context of static linear panel data models with interactive effects present in the error term and the regressors. Instruments are obtained from transformed regressors, thereby it is not necessary to search for external instruments. We consider a two-stage IV (2SIV) and a mean-group IV (MGIV) estimator for homogeneous and heterogeneous slope models, respectively. The asymptotic analysis reveals that: (i) the NT−−−√-consistent 2SIV estimator is free from asymptotic bias that may arise due to the estimation error of the interactive effects, while (ii) existing estimators can suffer from asymptotic bias; (iii) the proposed 2SIV estimator is asymptotically as efficient as existing estimators that eliminate interactive effects jointly in the regressors and the error, while (iv) the relative efficiency of the estimators that eliminate interactive effects only in the error term is indeterminate. A Monte Carlo study confirms good approximation quality of our asymptotic results.
We propose a new methodology for identifying and estimating demand in differentiated products models when demand and cost data are available. The method deals with the endogeneity of prices to demand shocks and the endogeneity of outputs to cost shocks by using cost data rather than instruments. Further, we allow for unobserved market size. Using Monte Carlo experiments, we show that our method works well in contexts where commonly used instruments are invalid. We also apply our method to the estimation of deposit demand in the US banking industry.
Juodis, Artūras & Sarafidis, Vasilis (2022)
A Linear Estimator for Factor-Augmented Fixed-T Panels With Endogenous Regressors
A novel method-of-moments approach is proposed for the estimation of factor-augmented panel data models with endogenous regressors when T is fixed. The underlying methodology involves approximating the unobserved common factors using observed factor proxies. The resulting moment conditions are linear in the parameters. The proposed approach addresses several issues which arise with existing nonlinear estimators that are available in fixed T panels, such as local minima-related problems, a sensitivity to particular normalization schemes, and a potential lack of global identification. We apply our approach to a large panel of households and estimate the price elasticity of urban water demand. A simulation study confirms that our approach performs well in finite samples.
Athanasopoulos, George; Sarafidis, Vasilis, Weatherburn, Don & Miller, Rohan (2021)
Longer-term impacts of trading restrictions on alcohol-related violence: insights from New South Wales, Australia
Background and Aims:In February 2014, the government of New South Wales (NSW),Australia, introduced new restrictions (known as the‘lockout laws’) on the sale of alco-hol in licensed premises in two of Sydney’s most prominent entertainment districts,Kings Cross (KX) and the central business district (CBD). This study aimed to determine:(i) whether the introduction of the lockout laws was the point at which the time patternof the assault series in the KX and CBD entertainment precincts changed; (ii) whetherthe apparent reduction in assault in these precincts persists when we control for com-mon variations in assault across the entire state of NSW; (iii) whether the reduction inassault in the KX and CBD entertainment precincts resulted in a displacement of theassault problem into other areas; and (iv) whether there is a net reduction in assault aftertaking any spill-over or displacement effects into account.Design:Structural break analysis was used to determine the date at which the time pat-tern of assaults changed. Interrupted time series analysis with a rest-of-NSW comparatorwas used to assess the change in assault.Setting, cases and measurements:The monthly totals of incidents of non-domesticassaults reported to the NSW Police between January 2009 and March 2019 (n= 123).Findings:The structural break in assaults occurred in January 2014 rather than inFebruary 2014, when the lockout laws were introduced. The reduction in assault persistseven when we control for common influences across NSW as a whole. In particular, fromJanuary 2014 onwards, assaults fell immediately by 22% (a downward step) in KX (90%confidence interval [CI] = 15–28) and by 33% in the CBD (90% CI = 19–47). Assaultscontinued declining in KX (trend-break coefficient =−0.094, 90% CI =−0.192 to 0.005).The reduction in assault in the KX and CBD precincts is associated with a rise in assaultin areas surrounding these precincts. The net effect, nonetheless, remains a lower levelof assault. In particular, we estimate that the net reduction over the three areas com-bined was 1670 assaults (i.e. 27 per month).Conclusion:Some of the initial reduction in assault in KX and the CBD of Sydney, Australia,previously attributed to the February 2014 introduction of lockout laws may have been aresponse to publicity surrounding recent deaths connected with alcohol-related violence.
Sarafidis, Vasilis & Wansbeek, Tom (2021)
Celebrating 40 years of panel data analysis: Past, present and future
The present special issue features a collection of papers presented at the 2017 International Panel Data Conference, hosted by the University of Macedonia in Thessaloniki, Greece. The conference marked the 40th anniversary of the inaugural International Panel Data Conference, which was held in 1977 at INSEE in Paris, under the auspices of the French National Centre for Scientific Research. As a collection, the papers appearing in this special issue of the Journal of Econometrics continue to advance the analysis of panel data, and paint a state-of-the-art picture of the field.
Kripfganz, Sebastian & Sarafidis, Vasilis (2021)
Instrumental-variable estimation of large-T panel-data models with common factors
In this article, we introduce the xtivdfreg command, which implements a general instrumental-variables (IV) approach for fitting panel-data models with many time-series observations, T, and unobserved common factors or interactive effects, as developed by Norkute et al. (2021, Journal of Econometrics 220: 416–446) and Cui et al. (2020a, ISER Discussion Paper 1101). The underlying idea of this approach is to project out the common factors from exogenous covariates using principal-components analysis and to run IV regression in both of two stages, using defactored covariates as instruments. The resulting two-stage IV estimator is valid for models with homogeneous or heterogeneous slope coefficients and has several advantages relative to existing popular approaches. In addition, the xtivdfreg command extends the two-stage IV approach in two major ways. First, the algorithm accommodates estimation of unbalanced panels. Second, the algorithm permits a flexible specification of instruments. We show that when one imposes zero factors, the xtivdfreg command can replicate the results of the popular Stata ivregress command. Notably, unlike ivregress, xtivdfreg permits estimation of the two-way error-components paneldata model with heterogeneous slope coefficients.
Juodis, Artūras & Sarafidis, Vasilis (2021)
An incidental parameters free inference approach for panels with common shocks
This paper develops a new method for testing for Granger non-causality in panel data models with large cross-sectional (N) and time series (T) dimensions. The method is valid in models with homogeneous or heterogeneous coefficients. The novelty of the proposed approach lies in the fact that under the null hypothesis, the Granger-causation parameters are all equal to zero, and thus they are homogeneous. Therefore, we put forward a pooled least-squares (fixed effects type) estimator for these parameters only. Pooling over cross sections guarantees that the estimator has a NT−−−√ convergence rate. In order to account for the well-known “Nickell bias”, the approach makes use of the well-known Split Panel Jackknife method. Subsequently, a Wald test is proposed, which is based on the bias-corrected estimator. Finite-sample evidence shows that the resulting approach performs well in a variety of settings and outperforms existing procedures. Using a panel data set of 350 U.S. banks observed during 56 quarters, we test for Granger non-causality between banks’ profitability and cost efficiency.
This paper develops two instrumental variable (IV) estimators for dynamic panel data models with exogenous covariates and a multifactor error structure when both the cross-sectional and time series dimensions, and respectively, are large. The main idea is to project out the common factors from the exogenous covariates of the model, and to construct instruments based on defactored covariates. For models with homogeneous slope coefficients, we propose a two-step IV estimator. In the first step, the model is estimated consistently by employing defactored covariates as instruments. In the second step, the entire model is defactored based on estimated factors extracted from the residuals of the first-step estimation, after which an IV regression is implemented using the same instruments as in step one. For models with heterogeneous slope coefficients, we propose a mean-group-type estimator, which involves the averaging of first-step IV estimates of cross-section-specific slopes. The proposed estimators do not need to seek for instrumental variables outside the model. Furthermore, these estimators are linear, and therefore computationally robust and inexpensive. Notably, they require no bias correction. We investigate the finite sample performances of the proposed estimators and associated statistical tests, and the results show that the estimators and the tests perform well even for small N and T.