Scholars now have unprecedented access to data from authoritarian regimes, thanks to initiatives like the Open Government Partnership. This article argues that while open data offers new research opportunities, it also introduces significant challenges related to potential manipulation.

In Tanzania, we demonstrate these risks by comparing publicly available tax revenue data with verified internal figures. The public data appear designed to present a favorable picture of government performance but actually significantly understate opposition success and overall governmental capacity.

This case reveals how authoritarian leaders strategically manipulate newly accessible data:

• Public data on local taxes systematically underestimate actual revenues
• These discrepancies support the regime narrative while distorting research findings
• The political context shapes both what gets released and its accuracy

We conclude by offering a practical framework for researchers to detect such manipulation in open government datasets, helping maintain research integrity as more closed regimes adopt data transparency practices.