In this paper we show that the double Pareto lognormal (DPLN) parameterization provides an excellent fit to the overall US city size distribution, regardless of whether 'cities' are administratively defined Census places as in Eeckhout (2004) or economically defined area clusters as in Rozenfeld...

We analyze the first data set on consistently defined functional urban areas in Europe and compare the European to the US urban system. City sizes in Europe do not follow a power law: the largest cities are "too small" to follow Zipf's law.

We analyze the first data set on consistently defined functional urban areas in Europe and compare the European to the US urban system. City sizes in Europe do not follow a power law: the largest cities are "too small" to follow Zipf's law.

There has been vast interest in the distribution of city sizes in an economy, but this research has largely neglected that cities also differ along another fundamental dimension: age. Using novel data on the foundation dates of almost 8,000 American cities, we find that older cities in the US...

There has been vast interest in the distribution of city sizes in an economy, but this research has largely neglected that cities also differ along another fundamental dimension: age. Using novel data on the foundation dates of more than 10,000 American cities, we find that older cities in the...

Older cities in the US tend to be larger than younger ones. The distribution of city sizes is, therefore, systematically related to the country's city age distribution. We introduce endogenous city creation into a dynamic economic model of an urban system. All cities exhibit the same long-run...

Older cities in the US tend to be larger than younger ones. The distribution of city sizes is, therefore, systematically related to the country's city age distribution. We introduce endogenous city creation into a dynamic economic model of an urban system. All cities exhibit the same long-run...

In this paper we show that the double Pareto lognormal (DPLN) parameterization provides an excellent fit to the overall US city size distribution, regardless of whether “cities” are administratively defined Census places or economically defined area clusters. We then consider an economic...

In this paper we show that the double Pareto lognormal (DPLN) parameterization provides an excellent fit to the overall US city size distribution, regardless of whether "cities" are administratively defined Census places or economically defined area clusters. We then consider an economic model...