Expectations about the future shape the way we make decisions. Economics has long understood the importance of expectations in deciding economic and social outcomes. However, evidence about the impact of expectations on non-monetary outcomes has been elusive and hard to identify.
In BSE Working Paper 1385, “Partisan Abortions,” Libertad González, Luis Guirola, and Blanca Zapater use unexpected electoral outcomes as a shock to identify patterns in fertility choices caused by expectational changes.
Partisanship determines how people react to events
People with different political alignments tend to adjust their expectations about the future in different ways after the results of an election. For instance, I might expect better conditions in the future for the economy if my preferred political party wins the general election. The opposite is true for a person supporting the losing party.
Although surveys have found an effect of political outcomes on peoples’ expectations, the evidence that the agents in fact act upon these expectational changes has been scarce. The authors use the case of the Spanish General Election of 2004 as a natural experiment to study how a plausibly unexpected outcome shapes individuals’ choices about conceptions and abortions in Spanish municipalities.
When the tables turn very fast
In the weeks preceding the Spanish General Election in 2004, a victory for the Conservative Party (the incumbent) seemed almost certain. However, an unexpected event, just days away from the election, disrupted everyone’s expectations. This was the case of the terrorist attacks in Madrid, which happened three days before election day, and were perceived to be badly managed by the Conservative Party then in power.
The events that followed took away from the Conservative Party a victory that seemed certain just days ago, handing the government to their rivals: the Social Democrat Party. Before the election, conservative voters had better expectations about the future of the economy when compared to social democrat voters. This shifted right around the time the election took place, as shown in Figure 1.
Changes in expectations affect fertility choices
These sudden changes triggered by the unexpected result of the 2004 election constitute a credible random shock to agents’ expectations, which is crucial for the credible identification of what drives the outcomes the authors want to explore. To be confident about how accurate voters are in forecasting electoral results, they show that in other elections polls accurately predict the result of the election.
Using administrative Spanish data at the municipality level about fertility and demographic characteristics for women in Spain between 18 and 44 years of age, the authors find a significant change in abortions and conceptions during the weeks following the election.
The authors find, through a difference-in-differences design, that abortion rates increased more for women in conservative-leaning municipalities when compared to those living in left-leaning municipalities, a spike of 18% relative to the mean. Moreover, they also find that the negative change in expectations for conservative voters led to a decrease in conceptions in conservative-leaning municipalities when compared to left-leaning ones.
The effect on abortions is mostly concentrated in the first few weeks after the election, while the effect on conceptions seems to be a little more persistent through time. All in all, the results points to an important effect of expectations on fertility choices.
The partisan divide has real implications
The authors make sure through a series of robustness tests that the terrorist attacks of May 2004 are not the source driving the changes in expectations and thus in fertility choices across municipalities. First, the results are not driven by municipalities in Madrid closer to the attacks, nor did the attacks seem to have affected differently the opinions of voters across partisan lines. Finally, they show that fertility choices also respond to outcomes of other elections.
Since the changes in fertility decisions take place immediately after the election results were out, and before the new government could implement any policy changes, the results suggest that shocks to economic expectations can affect fertility decisions.
More importantly, the reaction to such news is different across the political spectrum. Hence, the political divide seems to go beyond just public opinion matters, going all the way to different reactions by citizens to political outcomes in their day-to-day decisions.