Changes in the distribution of rental prices in Berlin

• I analyze changes in unconditional rental price distribution from 2007 to 2012 in Berlin.
• Rents are analyzed using quantile regression and decomposition methods.
• Prices increased throughout the entire distribution, but more at higher quantiles.
• The shift in unconditional rental prices is mainly due to altered coefficients.
• Changing characteristics have only little impact.

In this paper, I analyze the distribution of rental prices in Berlin using quantile estimates and decomposition methods. These methods have been rarely applied in the field of housing economics but have proven to offer relevant insights into the evolution of price distributions. The shift of rental prices from 2007 to 2012 is split into portions caused by changes in the distributions of the explanatory variables and by changes in their coefficients over time. Two main results stand out: firstly, quantile estimates illustrate the constraint of a mean regression as most coefficients differ substantially between quantiles; secondly, coefficients are the primary source of temporal difference in the distribution of rental prices. On the other hand, changing properties of flats being offered affect the increase in the rental price distribution relatively little.