This week, Economic & Political Weekly publishes the first thorough empirical application of my namematching algorithm: an exploration of the spatial variation of the "Muslim vote" in Gujarat and Uttar Pradesh in the 2014 general elections, which I have written with Raheel Dhattiwala (from the University of South Australia, and formerly Oxford), following up on Raheel's considerable research on Gujarat and my own earlier work on elections in UP. A postprint of our article is archived HERE, with the original version on the journal website:

Raphael Susewind, & Raheel Dhattiwala (2014). Spatial Variation in the 'Muslim Vote' in Gujarat and Uttar Pradesh, 2014 Economic & Political Weekly, 49 (39), 99-110

Since it has been quite a challenge for us to present a highly complex picture in one short overview article, we decided to supplement the print version with two interactive maps of our results. These show coverage, explanatory power and coefficients of our main SUR model across assembly segments; you can select your variable of interest in the bottom right corner; if you zoom in and mouse over an assembly segment, relevant details will pop uo in the top right corner:

Also, in line with my earlier endorsement of open data, we decided to make raw replication data and statistical scripts available under a Creative Commons CC-BY-NC-SA 4.0 license. The license implies that if you use this data in your own work, you have to attribute it (reference to either this dataset or the original article should do), can not use it for commercial purposes (academic publications are generally fine), and if you combine it with additional data, you have to make this additional data available to the general public on equal license terms.

The dataset, which contains among other things booth-level estimates of the Muslim electorate in both states, is archived HERE, thanks to the amazing data services unit at my university; as soon as I find time, I shall also update my own dedicated portal for data on religion and politics in UP.

Finally, as Raheel and I wrote in our conclusion, we hope that making these visualizations and raw data available leads to further and more detailed inquiry, replication and critique of our results, and ultimately the growth of booth-level analysis in Indian political science. This is just a beginning...

We gratefully acknowledge the use of the Oxford Advanced Research Computing unit (ARC) in carrying out this work.