Visualizing Telecom Inequality
Using historical data to tell the story of empire, colonialism, asymmetrical access to information
Data-gathering has long been a crucial part of governing global communications networks.
From the earliest days of electronic communications, national and colonial administrations gathered statistics to keep track of where and how global networks were growing.
These statistics tabulated all kinds of things: the number of miles of telegraph cable laid, on land and under the sea; the number of telegraphs delivered over those cables, separated out between those for public use, those for official use, and those used by the press industry; the number of radio stations and transmitters installed; even the number of employees working in the public telecom system, separated out between “foreign” and “local.”
Nation-states and colonial administrations collected this data for their own administrative purposes. But, beginning in the 1860s, they began to collate and share this data through the international organization founded to help govern these networks: the International Telecommunication Union. And, starting in 1865, the ITU began to publish these statistics for public consultation every single year— something the organization continues to do to this day.
Here’s what these statistics look like
These statistics are a treasure trove to anyone interested in the history of empire, colonialism, and infrastructures of globalization in the 19th and 20th centuries.
There’s just one problem: they have yet to be digitized.
ITU archivists have done a tremendous public service by uploading scans of these tables to their online archives. But the statistics have yet to be converted into digital tables for data analysis. That is the work that my students and I have been doing, incrementally, over the last couple of years.
As we digitize this data, we are learning fascinating insights about the history of communications and colonialism. And, progressively, we are able to create heat maps of the flow of information across borders from the height to the collapse of colonial empires.
Here, for example, is what telegraph statistics for the year 1934 look like when digitized
This data set is full of potential, but there are a couple “problems.”
These “problems” make the data set challenging to work with— but also point toward even more fascinating historical questions.
First: the names of the countries and colonies change several times over the course of this century-long data set.
This shouldn’t surprise us, given how imperial expansion and decolonization produce chaotic and shifting borders.
But it makes digitization tricky, and transposition to digital mapping even harder— because it can make it difficult to know what the geographic referent is for countries and colonies listed.
It’s particularly difficult for data that refers to colonies in Africa, Asia, and the Caribbean, where colonial borders and administration practices could change rapidly and often.
Second: the data schema changes several times over the data set.
That is to say— from year to year, the data set quantifies different things.
Every year, officials of the International Telecommunication Union (ITU) sent out identical questionnaires for its nation-state members to fill out and send back. But at least five times in this 100-year stretch, ITU officials changed the prompts they sent to national administrations.
They did this both because technologies were evolving, and could do new/different things they wanted to tabulate. But the prompts also changed as the kinds of information that engineers and administrators found valuable, or wanted to know, evolved. Sometimes a prompt would disappear from the questionnaire, or it would be slightly revised or retooled.
This says something interesting about the changing priorities of international administration and scientific cooperation. But it can make it difficult to track the exact same data point over the whole data set. Some data prompts/points that do remain consistent, however, include the number of miles of telegraph cable in operation, and the volume of total public telegraphs sent and received.
Example of one page of “footnotes” to the data
Third: the data’s availability is inconsistent.
Submitting data was entirely voluntary for ITU member-states, and some administrations submitted their data only a handful of times in the 120-year run.
Submitting data was entirely voluntary for ITU member-states, and some states submitted data only a handful of times in the 120-year run. In other cases, states did submit data, but either didn’t have on hand or weren’t interested in supplying answers for every single prompt.
The data set is also full of footnotes and qualifiers submitted by officials, which contextualize, explain, or in some ways amend the otherwise sterile numbers written in the table.
These footnotes are catnip as a historical trace, but they also make the data somewhat resistant to the boxes of a digital table.
Here are some examples of insights this data set can provide
The following visualizations are made from data listed in tables on international telegraphy from the year 1934, collated from the section in red below.
These categories tabulate the total amount of telegraph traffic sent over a country or colony’s domestic network, divided between the location of the telegram’s final destination.
Specifically, they split telegraph traffic sent exclusively within the country itself; (non-domestic) traffic within the same continent; traffic with Europe; and traffic with other continents.
This data, in other words, can give a loose sense of the volume of communication that each country or colony shared with other countries— how connected they were with their continental neighbors, with their (European) metropole, and/or other continents.
We can start with some simple bar charts.
