We have been creating high-resolution GIS maps of the boundaries of electoral circumscriptions (a.k.a. electoral constituencies or districts) used in Lower Chamber elections all over the world. Currently, we cover 1341 elections in 164 countries (>25Gb) and an accompanying paper is under works.

Below you can have a sneak peek at the world’s election boundaries circa 2020:

All boundaries have been recreated following strictly official information, legislation and/or official map images. The vast majority are high-resolution boundaries created starting from official GIS administrative or census maps, and carefully manually adjusted or vectorized as needed. Even in the case of historical boundaries, we have been making great efforts to re-utilize parts of more current official boundaries and manually alter only what is required to account for the over-time changes.

For just 10 countries we can only roughly approximate boundaries. And we already have materials to improve some of those, like India, given proper time and resources. Only 25 countries are still entirely missing – if we count as Not Applicable those that either never had any Lower Chamber election (e.g. Oman, Saudi Arabia, uninhabited areas), never had one with more than a single contesting party/group/individual (e.g. Eritrea, North Korea), did not have one in more than 50 years (e.g. Cuba, Laos) or had some but not using geographically defined electoral circumscriptions (e.g. Libya, Sudan).

All that said, note that the above visualization is still a work-in-progress that needs cleaning. Also, keep in mind that boundaries between countries often do not have precise matches precisely because each official maps and data may have different views about the boundaries, different accuracy and different resolution.

If you are curious to learn what exact elections from which countries we already have in my dataset, here is a list:

This project began many years ago as the GeoReferenced Electoral Districts Datasets (GRED) , which we have been leading at the Constituency-Level Elections Archive (CLEA) project hosted by the Institute for Social Research, University of Michigan. GRED has a a bit more than 100 maps (and should get a new release very soon!), targeted at elections currently present in CLEA's covered election results; it mostly focuses on countries' current elections only; they are not necessarily high-resolution maps; and they are mostly not created from official sources (for licensing easiness).

My dataset is more ambitious than GRED both in terms of coverage and of quality, and it also includes elections not yet available in CLEA or else where. However, creating high-resolution maps based on official data only, imposes a trade-off cost: maps in my dataset cannot be all released under the same usage licences. This happens because some data used as raw materials to create some of the boundaries are available under a "share-alike" licence - which means that they require that derivative projects follow their same licensing model. But it turns out those are many! So, together with detailed explanations on how each boundary map was created, the data documentation will inform the licence of each election map individually. However, fear not: whatever each licence, they will all be free to use for any non-commercial purpose!

A huge thanks must go to all CLEA team, without whom none of those projects would have been possible. We are particularly thankful to Allen Hicken and Ken Kollman, for their advising, guidance and trust, and to Charles Zinn for the amazing assistanship. Charles is a tremendously talented young researcher, extremely skilled with geospatial data and, in particular, vectorization. He has helped me a lot in creating the upcoming version of GRED and my interactions with him made me learn quite a bit of the technical problems we needed to overcome for this new dataset we are creating now. Finally, we are also grateful to Julia Lippman and Yioryos Nardis for present and past tremendously helpful administrative support.

This project would also not have been possible without the kind help of hundreds of people all around the world that we have contacted for almost a decade (even before coming to Umich!) while collecting the materials used in those maps. We are trying hard to name them all in the upcoming documentation of my data.