What about that dowry inflation in South Asia?

In a National Bureau of Economic Research paper Raj Arunachalam and Trevon Logan say it's a myth and theoretical papers have been trying to explain something for which there is no systematic evidence:

Using large-sample retrospective survey data from India, Bangladesh, Pakistan, and Nepal, we assess the empirical evidence for dowry inflation. We find no evidence that real dowry amounts have systematically increased over time in South Asia.

The same question could be asked about dowries in the UAE. The claim is that they are 'spiraling out of control' (as it's often put), but is that really true?

Here is an ungated version of the paper.

Thanks to Chris Blattman for the pointer.