Ahmed Sadek Yousuf

Post Doctoral Research Fellow, Aix Marseille School of Economics

University of Arkansas

Email: ahmed-sadek.YOUSUF@univ-amu.fr; asyousuf@outlook.com

CV (Updated: August 2022)

Research Statement (Updated: September 2022)

Currently a Postdoctoral Research Fellow at Aix Marseille School of Economics under the TMENA2 project to empirically ascertain conflict and socioeconomic dynamics as well as rivers of radicalization in Middle East and Europe. Obtained PhD Economics from University of Arkansas in 2021 July. My research is primarily concerned with application of advanced spatial econometric methodologies as well as satellite derived data in investigating the intersection between development economics and spatial / regional economics, as well as economic growth related concerns, and also looking into the broader dynamics of conflict under the broad umbrella of Spatial Economics.

Research

Working Papers & Works in Progress:

Estimation of Count Data Models with Multiplicative Fixed Effects and Spatial Dependence : This note looks at estimation of spatial autoregressive models for non-negative and count outcomes with multiplicative fixed effects. We show that in presence of significant proportion of zeros in the count variable, control function and Instrumental Variable estimation to model such spatial dependence yields biased and inconsistent estimates, which can be primarily attributed to the assumed form of proxy for the spatial autoregressive term. This may be remedied by using a different form of proxy. We also show that consistent estimates can be derived with Conditional Poisson Maximum Likelihood based on modified `reduced' form conditional mean specification. The properties of these estimators are subsequently investigated in Monte Carlo trials. Submitted at Economic Letters

(Link: https://drive.google.com/file/d/1bB4xbK5jJ5NLK17hnumCe4_cwkaca005/view?usp=share_link)

Daily Night Lights and Conflict in Bangladesh} : (With Dr. Christophe Muller at Aix Marseille School of Economics): Leveraging new daily night lights data (sourced from NASA), alongside Dr. Christophe Muller, we investigate how political violence (and its various definitions, such as hartals) affect (daily) economic activity in Bangladesh at (sub-district level). Here we are looking to identify short term effects of for example hartals (in the context of Dhaka, it may be seen as a dip in regular activity just after a political violence / hartal incidence and therefore very likely to be picked up by night lights). Given that the daily night luminosity data is recorded at 1:30 am local time, thus the main explanatory variable (conflict event count) is lagged by one day. The methodological challenges came from accommodating the non stationary nature of such high frequency night lights data which would have made ordinary inferences (using conventional inference tools like Ordinary Least Squares) quite difficult. Thus time series ARCH (Autoregressive Conditional Heteroskedasticity) models were used to model these interactions (between night lights and political violence). Political Violence data is derived from ACLED (Armed Conflict Location and Event Data project). From the dataset, we use keywords to further derive statistics for different definitions of violent incident (such as hartal), not otherwise included in the ACLED description. As such, most prominently, we already find a statistically significant effect of -0.9 percent immediate effect of hartals (aka political strikes) on daily night lights in Dhaka city for example. This paper differs in that in the conflict literature by and far yearly data was used, and this is the first instance of using such high frequency data to identify immediate/ short term effects.

A Spatial Poisson 'Full information' Fixed Effects, Dynamic Panel Estimation Approach : I present Maximum likelihood estimation framework of Exponential Feedback Models with multiplicative spatial spatial autoregressive laga and fixed effects, with Monte Carlo results showing significant improvements over alternative approaches such as Generalized Method of Moments estimators. When a lagged dependent variable (in log form) is included in the conditional mean specification, the consequent estimator inconsistency is corrected by means of Split / Half Panel Jackknife estimation method. Two applications are presented regarding dynamic spatial count model: one using the dataset on Part 1 and Part 2 Crimes in Pittsburgh from Liesenfeld (2016) and secondly leveraging the Los Angeles County historical dataset on Part 1 and Part 2 crimes, to investigate the explanatory and predictive power of Part 2 crimes regarding Part 1 crimes (aka broken windows hypothesis). In addition, I also present a Spatial Fixed Effects Zero Inflated Poisson model approach, using a composite likelihood approach with again Monte Carlo results pointing to its consistency. The development of these models are meant to facilitate accurate modeling of dynamics of variables which has a count nature (such as number of crimes, conflicts, patent counts, etc) and may have an embedded spatial dimension aspect to them. With regards to Spatial Panel Zero Inflated Poisson application, I leverage the framework to look at impact of specific climate indicators on conflict in Africa (at grid level).

Night Lights and Conflict in Sub Saharan Africa: Adapted from my PhD Job Market Paper, here I construct a bi-variate Spatial Vector Autoregression Model and consequent Impulse Response framework, to investigate the dynamics between economic activity (as proxied by night lights) and log of conflict count in Sub-Saharan Africa at ADM1 / provincial level between 1997-2013. I further investigate the dynamics concerning 2 other definitions of conflict: peaceful protests and non-peaceful protest events. The primary equations are estimated with bias corrected Split-Panel Jackknife Spatial Maximum Likelihood, so as to correct the parameter inconsistency that results from inclusion of lagged dependent variable(s) in the specification, as opposed to the more common GMM (Generalized Method of Moments) methods for estimating such models prevalent in the literature. This in it of itself is an alternative strategy to model Panel Vector Autoregressive framrwork. The resulting direct, spillover and average total impulse responses are significant with the latter 2 showing a degree of persistence, and the impulse response standard errors are derived by a recursive wild bootstrap procedure. To be submitted to The Review of Economics and Statistics.

Deriving Economic Growth Estimates with 2 Satellite signals} : (With Andrea Civelli \& Aryah Gaduh of University of Arkansas). An extension of the third Chapter in my Dissertation, in collaboration with dissertation supervisor Andrea Civelli, with implementation being at the world / national level, and comparing the estimates to the original lights derived only economic growth estimates from Vernon Henderson (2012). The Night Lights data is now extended to utilize data to up to 2019/20 as well as including the more up-to-date VIIRS derived Night lights which improves over the original DMSP OLS night lights in a number of ways, such as top-censuring.

Half Panel Jackknife Estimation of Spatial Dynamic Panel Models with AR(1) Error terms and predetermined regressors} : In this study I am working on an estimation framework for a spatial dynamic model with an AR(1) error term as well as predetermined regressors which has an in-built strong level of correlation with lagged disturbances.The estimation framework is based on standard Spatial Maximum Likelihood estimation with either Prais Winsten or Cochraine Orcutt transformation to deal with the autocorellation, and subsequently implementing a Half-Panel Jackknife correction to correct the bias present on the lagged dependent variable. Furthermore, Forward Orthogonal Deviation transformation is applied to the data before estimation, instead of the more conventional fixed effects. Monte Carlo results show significant improvements compared to more conventional models which do not factor in the autocorrellation, something general spatial GMM (Generalized Method of Moments) approaches may not handle well.

Assessing Dynamics between World Bank Aid Allocation, Economic Growth and Conflict Intensity in Sub Saharan Africa using a Spatially augmented P-VAR}} (Job Market Paper): Using a Spatial Panel Vector Autoregressive Model, with spatial dependent disturbances, I explore the dynamics between the aforementioned 3 variables at provincial level in Sub Saharan Africa, estimated in a multi-stage non-linear Generalized Method of Moments approach. Following recent literature, I used night light emissions recorded by National Oceanic and Atmospheric Administration to derive regional level of economic growth estimates. Later, I decompose the impulse responses of the system into a direct and indirect (average spillover) components. The GMM approaches include the methodology as introduced by Lung Fei Lee and applied in a Continuously Updated Estimator Framework. Furthermore, leveraging the spatial structure of the estimation framework, both direct and spillover impulse responses were derived. (Link)

Conflict Intensity and Inequality in Africa: Presenting an Empirical Likelihood based Spatial Count Model estimation procedure with multiplicative dynamic and spatial effects, I look at the effect of economic inequality on conflict incidence in Sub-Saharan Africa, at sub-national level. (Link )

Deriving Economic Growth Estimates with Multiple Satellite signals: Here I present a theoretical framework by which I estimate 'true' measure of economic growth / activity by augmenting Vernon Henderson's (2012) methodology (where he used night lights to measure economic growth at national level) with a second signal, at subnational level. I ran exercises on India, Brazil and Indonesia (using loss of vegetative cover as the second signal) and also for the US. I also present a Generalized Method of Moments estimation framework in the presence of more than two (independent) 'signals'.Link

Classes taught:

Fall 2017 : Econ 2143 Basic Economics: Theory and Practice (1 Full Section)
Spring 2018 : Econ 2143 Basic Economics: Theory and Practice (1 Full Section)
Fall 2018 : Econ 2023 Principles of Microeconomics (1 Full Section)
Spring 2019 : Econ 2023 Principles of Microeconomics (1 Full Section)
Fall 2019 : Econ 2023 Principles of Microeconomics (1 Full Section)
Spring 2020 : Econ 2023 Principles of Microeconomics (1 Full Section)
Fall 2020 : Econ 2023 Principles of Microeconomics (1 Full Section)
Spring 2021 : Econ 2023 Principles of Microeconomics (1 Full Section)

Teaching Assistant:

ECON 3033 Microeconomic Theory : Spring 2017; Fall 2018

ECON 3843 Economic Development, Poverty and Role of Development Institutions : Spring 2017; Fall 2018

ECON 3433 Money and Banking: Spring 2016; Spring 2017; Fall 2018

ECON 4643 International Macroeconomics and Finance: Spring 2016; Spring 2017; Fall 2018

ECON 4743 Introduction to Econometrics: Fall 2019; Spring 2020

Teaching

Page updated

Google Sites

Report abuse