Until now, research suggested that poorer voters wanted Brexit, no matter the costs, and richer voters chose Remain.
But there’s a lot of heterogeneity in Leave and Remain votes. And a substantial % of strong eurosceptics voted to Remain.
Project Fear didn’t work, right? We think it likely did.
We think of Brexit as a vote against a status-quo. This factors in the net benefits a voter may have from changing the status-quo. We need to think of risks net of costs-benefits calculations. We do that in our paper.
If wealth insures against risk, we would expect wealthier voters to be more likely to support changes to the status-quo (i.e. Leave), and poorer voters to be more likely to support Remain instead.
Empirically, we are able to test this thanks to access to *individual-level* wealth data, using @BESResearch data, @bankofengland panel data and a survey experiment in the BES.
Across all three analyses, we find wealth (esp. property) increases support for Leave. We find support for the proposed insurance mechanism:
(1) we show wealth impacts the voter’s expectations that Brexit would not have a detrimental effect on personal economic well-being, and
(2) we find that wealth leads to lower risk-aversion
We provide arguments for why property wealth should provide greater insurance in this context. Our results also hold over high/low aggregate wealth (house price) variation, and are stronger in the wealthiest areas.
Why it matters? Wealth insures against econ risks in political choice may apply in a wider range of electoral choices/SQ changes (e.g. secession decision, incumbent/challenger selection). Importantly, individual wealth likely matters more than aggregate levels of wealth.
Why does all this matter? (2) In the case of Brexit, our results show that there were poorer voters who felt they couldn’t support Brexit because of the potential impact on their economic circumstances...
... For them, Brexit’s success therefore has important political implications. At the other end, wealthier voters are expected to remain largely unaffected.