No abstract

Empirical evidence for the fitting of cities size of Romania both from models having Pareto tails and from models which do not, situations that occur at the same time  Comparisons of Pareto tails and log-normal or generalized beta of second kind (GB2) body distributions to mixtures of log-normal models  Statistical equivalence of mixture of log-normal models to Pareto tails and different bodies  Introduction of the threshold double Pareto GB2 model (tdPGB2) *Highlights (for review)

We study the parametric description of the city size distribution of four European countries: France, Germany, Italy and Spain. The parametric models used are the lognormal, the double Pareto lognormal, the normal-Box-Cox (defined in this paper) and the threshold double Pareto Singh-Maddala (introduced in a cited recent paper when studying US city size).The results are quite regular. The preferred model is always the threshold double Pareto Singh-Maddala in the four countries. However, the dPln is not rejected always for the case of France, and in the case of Italy the dPln is the runner-up distribution. These results complement those obtained in a cited recent paper which study the US places' city size distribution.JEL:C13, C16, R00.

We study the parametric description of the city size distribution (CSD) of 70 different countries (developed and developing) using seven models, as follows: the lognormal (LN), the loglogistic (LL), the double Pareto lognormal (dPLN), the two-lognormal (2LN), the two-loglogistic (2LL), the three-lognormal (3LN) and the three-loglogistic (3LL). Our results show that 3LN and 3LL are the best densities in terms of non-rejections out of stan-Financial support from Ministerio de Economía y Competitividad (ECO2017-82246-P) and support by Aragon Government (ADETRE Consolidated Group) is acknowledged.

No abstract