This report is particularly suspect. First of all zero-rating is not banned in the EU, and it's not clear which countries are included in the "has zero-rating" basket. The more likely interpretation of their data is that "in countries with shitty internet, providers tend to offer a lot of zero-rating offers". E.g the internet is vastly better in romania than in greece yet they both have a lot of "differentially rated" offers<p>Then they only show two years , 2015 vs 2016, where there is a slight increase of 2% in prices , without error bars. Then there is this:<p>> we repeated our analysis for zero-rating offers introduced in 2016 or 2017. However, initially this did not produce statistically signifcant results in any category. Closer examination of the data however revealed Finland to be an outlier market, in which the replacement of a single offer signifcantly changed the prices in almost all data volume aaskets. This is likely due to the fact that unlimited data plans, which do not sensibly admit a price per gigaayte calculation, are prevalent in Finland. We therefore repeated the analysis but excluded Finland from our dataset. In this case, we found a statistically signifcant result (p=0.04) for markets in which zero-rating was introduced between 2015 and 2016. These markets showed a 1% price increase between 2016 and 2017, whereas markets without zero-rating in both cases showed a 10% price decrease.<p>I think they are stretching it with p=0.04 on a cherrypicked sample of n=30, and present a rather peculiar conclusion about their data. Zero rating is obviously marketing garbage, but i am very unconvinced that it is the reason why ISPs are not investing in their networks.<p>(It also took 10 minutes to download their 5MB pdf - talk about bad internet ;) )