A hundred year journey
This month India hit a historic milestone. One hundred years ago, in November 1920, Indians in direct-rule British provinces, albeit a small section of the elite, went to the polls for the very first time and elected their representatives. Following the elections, the first provincial councils with majority Indians sat down to legislative business in January 2021.
The elections took place three years after an announcement by the Secretary of State of India, Edwin Montagu, in the English House of Commons in August 1917. In the British Crown’s view the reforms announcement of 1917 was:
“… to be the most momentous utterance ever made in India’s chequered history. They pledge the British Government in the clearest terms to the adoption of a new policy towards three hundred millions of people” (Report on Indian Constitutional Reforms, page 1, 1918).
The franchise extension was based on male property qualifications in each province, defined as payment of land revenue, land rent in rural areas, local municipal rates in urban areas, professional tax or income tax. This resulted in anywhere between 1.2 to 5 percent of a province gaining the right to vote (see Table 1 below). A majority of the provinces adopted first-past-the-post voting rules as the simplest method of elections given the relative inexperience of voters. Only the Madras and Bombay Presidencies adopted proportional representation voting rules to accommodate local caste dynamics. Colored ballot boxes were used to represent candidates in order to accommodate illiterate voters.

In an attempt to increase the “association of Indians in every branch of the administration, and the general development of self-government institutions, with a view to the progressive realization of responsible government in India”, the Crown instituted a series of political, bureaucratic, and fiscal reforms. These resulted in a significant devolution of powers and responsibilities to the Imperial and Provincial Councils of India.
The elections introduced many of contemporary India’s electoral peculiarities such as a robust state role in elections administration, a system for contesting corrupt elections, and of course, the current electoral system. The header picture in this post shows members of the Justice Party in the Madras Presidency in the 1920s. This party would eventually transform into one of the main political parties in the state of Tamilnadu in the post independence period — the DMK. While historians have long noted the significance of these 1920 elections for the local politics of regions as different as Madras, Bombay, and Bengal, their collective weight suggests that these reforms were no more than hand waving by the British, and as Gilmartin and Moog note “instituted by the regime as much to bolster its own authority (by appealing to a conservative rural voice).” That impression might now be changing.
India’s colonial-era experience of franchise expansion presents an interesting opportunity for social scientists. India joins other colonies such as Canada, Senegal, Hong Kong, Sri Lanka and many others that experienced democratic institutions before full democracy.
In a previous post I noted how much of the theoretical thinking on colonial legacies has focused on geographic or early economic institutions. Research on colonial legacies has largely ignored the impact of early democratic institutions on post-colonial politics or institutional strength.
Research in many developed countries has found that early democratic institutions in fact, do matter. Take for instance Mares and Queralt’s paper on how landed elites in the 19th century in many Western countries were more likely to support the creation of an income tax if franchise extension was low. This was in large part because they could shift the burden of future taxation on to industrial elites in anticipation of redistributive taxation after an even greater franchise expansion. Separately, Ansell and Samuels find that industrial elites in contexts of higher income inequality were more likely to support democratization in order to seek more credible protection from state expropriation (see also related findings by Lizzeri and Persico). Others, like Berlinski and Dewan studying franchise expansion in 1867 United Kingdom, have identified the political effects of franchise extension by exploiting sharp changes to the electorate, and shown that increasing the electorate make constituencies more competitive.
Based on the findings in the West it is likely that colonial era democratic institutions shaped not only political competition, democratic consolidation and electoral norms, but also the character and substance of state policies and state capacity. But before we rush to apply some of the lessons from limited franchise expansion in the West to India, it might be worth considering how colonialism shaped elite expectations of politics in this era. What issues animated the first elections and how might they differ from the political battles between say aristocrats and capitalists in early 19th century England?
Studies that focus on the decade preceding the elections offer up some clues. Latika Chaudary for instance shows how upper castes wielded disproportionate control over educational spending, a key public good provided by the colonial state. It is therefore likely that education policy emerged as a flash point between upper and middle castes in the early legislatures. Guo Xu’s recent working paper on the role of Indians in the British bureaucracy in fighting the 1918 flu suggests that Indians understood the value of controlling and directing state resources, which they would be able to do with revenues raised from land taxes in the newly elected provincial councils. Banerjee and Iyer’s study of early colonial economic institutions suggests that elites from landlord versus non-landlord districts likely had differing preferences for the routine upkeep and implementation of revenue settlements surveys and tax machinery upgrades that the council would now control. Each of these suggest different predictions for the political parties and coalitions that were to emerge.
So what do we know about first elections?
Of the 5.5 million eligible voters, only 1.5 million voters actually voted. In a new project, my co-authors Ashish Aggarwal, Ritam Chaurey and I explore the determinants of voting and competition. We find that districts with greater caste fragmentation- measured using the 1920 district-level census proportions of Advanced castes, Intermediate castes, Backwards castes, Depressed Classes, Tribes, Muslims and others — had a higher level of turnout, more closely contested elections, and less likelihood that a candidate ran uncontested (see figures 1 and 2 that are bin scatter plots of data from 146 districts below). We find no similar effects for subcaste fragmentation. Our paper shows not only that ethnicity emerged as salient in politics in the first elections, but that status caste competition between upper and middle castes seemed to have mattered more than simply ethnic heterogeneity measures using jati (subcastes) fragmentation.


But how do these early elections matter beyond the elite politics of 1920? In a working paper, Cassan, Iyer, and Mirza examine elections between 1920 to 1957 and track how multiple rounds of franchise expansion changed competitive dynamics. They find that when franchise was expanded after 1935 and then again after 1950, the extent to which turnout increased was substantially smaller than the increase in enfranchisement. Relatedly the number of candidates standing for elections increased after every round of reforms but this increase was at a much slower pace than the franchise expansion.
Finally, there is one other reason why India’s elections might be of interest to scholars. Franchise expansion to the provinces of colonial India arrived somewhat exogenously. While the British were concerned about a growing nationalist movement in India and its role in World War 1, changing sentiments amongst the British population and the pivotal role played by India’s Secretary of State Edwin Montagu himself in championing the reforms precipitated the events I describe above. Unlike a considerable literature in political economy that focuses on how inequality, taxation, or conflict might shape democratization, India’s experience allows us to study the reverse — how did franchise expansion shape spending priorities, taxation, and inequality? These are burgeoning areas of interest to the field, and I hope to revisit these themes in another post soon when we have more evidence!