It’s often said that new leaders experience political honeymoons during their first days in office, when they enjoy the most latitude to change public policy. A number of self-styled reformers have recently risen to power, from Narendra Modi in India to Emmanuel Macron in France, and each time, analysts have highlighted the good will surrounding these leaders in their first terms.1,2 Less clear is how long their honeymoons tend to last.

In theory, public opinion polls are the best source of information about leaders’ favorability, but they are too infrequent and methodologically inconsistent to use in studying political honeymoons, and are not available for many countries. To work around this data gap, we can also gauge honeymoon periods through sentiment analysis of news stories. In particular, we can identify the set of articles about a given leader each day; compute the articles’ net level of positive emotion using one of several common sentiment lexicons; and then chart this daily average sentiment as a function of days since the leader assumed office. This simple metric gives us a rough picture of how a leader's image in the press changes during her early years in office, which is a decent proxy for the extent of a political honeymoon. Moreover, the computational consistency and near-real-time frequency of the measure allows for some interesting comparisons.

The plot below shows the net positive sentiment in news each day for a select group of leaders, starting from their respective first days in office.2 Whether sentiment scores are positive or negative per se matters less here, as baseline sentiment scores will vary depending upon the sentiment lexicon one chooses for the analysis. Instead, it is the trajectory of each leader over time, and the comparisons of scores across leaders, that provide the most insight, and these stay largely consistent no matter which sentiment lexicon one uses.

v6_blog

One interesting feature of the chart is the clear difference in starting points across leaders, likely accounted for by domestic politics. Both Jae-In and Ramaphosa assumed the presidency under highly irregular circumstances, as their predecessors either resigned or were impeached over corruption charges. By our metric, initial feelings about them were much less positive than for leaders sworn in after regular, democratic elections.

The most striking aspect of the graph, however, is the lack of consistent honeymoon patterns, particularly in leaders’ first hundred days. In fact, Widodo is the only one whose overall positive sentiment monotonically decreases in line with a honeymoon story, where good will starts high but is progressively exhausted. Ramaphosa’s trajectory is relatively flat; Macron and Modi’s scores make a relatively quick U-turn; and Jae-In’s actually increases. Honeymoon effects remain inconsistent across the group when we consider a longer time horizon, with only Widodo and Macron’s sentiment scores having fallen considerably below their starting points one year after assuming office. While it’s tempting to paint each new reformer on the block as a honeymooner, political realities are more complex than this simple trope allows.

Jason Weinreb

Quantitative Geopolitical Analyst at Koto, political science Ph.D. from Stanford, former analyst at Variant Perception Macroeconomic Research.