1066 and all that: Some deep determinants of voting shares in the 2016 referendum on EU membership

• DOI: http://doi.org/10.1111/twec.12621
• Date: 01 April 2018
• Published date: 01 April 2018
• Author: David Fielding

SPECIAL ISSUE ARTICLE

1066 and all that: Some deep determinants of voting

shares in the 2016 referendum on EU membership

David Fielding

University of Otago, Dunedin, New Zealand

1

|

INTRODUCTION

The Department of Economics at the University of Nottingham has established a reputation as a

world-leading centre for scholarship in the economics of international trade, in large part through

the leadership of Chris Milner and David Greenaway over the last 20–30 years. In addition to the

large volume of high-impact research, Nottingham has educated generations of undergraduates in

trade theory and its application, so that the average Nottingham economics graduate is well able to

understand the complex issues relating to international trade policy and UK membersh ip of the

European Union. However, these issues appear not to have been to the fore during the 2016 refer-

endum on EU membership: the Remain campaign did not try very hard to educate people about

comparative advantage, the Leave campaign spent little time discussing the distributional implica-

tions of the Stolper–Samuelson theorem, and econometric estimates of trade creation and trade

diversion effects have not featured prominently in recent parliamentary debates about membership

of the EU customs union. At the end of this paper, I will discuss some of the implic ations of this

disconnect for economics departments such as Nottingham’s, but this discussion will be informed

by an analysis of some of the factors that do explain voting in the referendum.

1

For reasons

explained below, this analysis will focus on England rather than the Celtic nations and on the

“deep determinants”of voter choice embedded in English social and economic history.

2

|

EU MEMBERSHIP, LIBERAL VALUES AND ENGLISH

HISTORY

Fielding (2017a) presents an analysis of some of the long-run historical factors that explain regio-

nal variation in the attitudes expressed in surveys such as the British Election Study (BES). It

appears that in the twenty-first century, inhabitants of locations showing evidence of exposure to

medieval ethnic and religious diversity are significantly more likely to express positive views about

immigration and equal rights for minority groups.

2

One possible explanation is that the initial

1

The analysis in this paper is informed by Fielding (2017b), which includes a brief discussion about attitudes towards EU

membership but no discussion or analysis of voting patterns in the 2016 referendum.

2

Other papers which explore the historical origins of regional variation in modern attitudes include Jha (2013), Voigtl€

ander

and Voth (2013) and Alesina, Giuliano, and Nunn (2013).

DOI: 10.1111/twec.12621

World Econ. 2018;41:1131–1146. wileyonlinelibrary.com/journal/twec ©2018 John Wiley & Sons Ltd

|

1131

exposure weakened prejudice towards other groups and that this new cultural norm was transmitted

to subsequent generations. The exposure effect is consistent with findings in social psychology

which suggest that under certain conditions—including equal social status and an absence of direct

competition—personal contact with members of another group can have a positive effect on atti-

tudes towards them.

3

Moreover, the formal theoretical model of Cavalli-Sforza and Feldman

(1973) implies that intergenerational transmission through socialisation of the young could lead to

persistent regional variation if the socialisation takes the form of “many-to-one”interactions—that

is, the whole village raising the child—see Cavalli-Sforza (1981) for more detail. Evidence pre-

sented in Fielding (2017b) suggests that at least part of the intergenerational transmission has be en

through an educational channel: ceteris paribus, locations with more exposure to medieval diver sity

had a significantly greater density of educational institutions during the Enlightenment, suggestin g

that diversity was associated with a taste for new ideas. The regional variation in the density of

these early institutions is strongly correlated with variation in the density of modern univer sities,

and inhabitants of modern university towns are significantly more likely to express positive views

about immigration and equal rights than are inhabitants of other towns. This is partly because

inhabitants of university towns have markedly different personal characteristics (e.g. they are

younger, on average, and have better educational qualifications), but a significant effect remains

even when we control for these characteristics. In other words, we might expect the average taxi

driver in a university town to have views on immigration and minority rights that are significantly

more positive than those of taxi drivers elsewhere.

There is evidence that opinions about the UK’s membership of the EU are also strongly corre-

lated with an individual’s education level (e.g., Bruter, 2005: appendix 3) and one explanation for

this similarity is that feelings about EU membership, immigration and minority rights all reflect

one’s position on a socially liberal/socially conservative spectrum, this position depending partly

on one’s level of education. In this case, it is possible that voting patterns in the 2016 referendum

reflect the same deep historical determinants that have been shown to explain some of the variation

in attitudes towards immigration and minority rights. Before discussing the data and evidence relat-

ing to voting patterns, we provide some context by briefly reviewing the correlates of exposure to

medieval diversity and early educational institutions introduced in previous papers.

2.1

|

Medieval diversity #1: Jewish settlement

The one large immigrant community in early medieval England were the Jews. The first Jewish

immigrants arrived from France at the end of the eleventh century, and tax records indicate that by

the end of the twelfth century, Jewish communities had been established in about 20 English towns

(Hillaby, 2003). Jewish settlement appears to have been encouraged by Norman and Angevin

kings, and the Jews fulfilled two main economic functions (Mundill, 2010): the provision of finan -

cial services at a time when Christians were forbidden to lend money to each other at interest, and

the creation of a tax base. Unlike any other commoners, the Jews were vassals of the king, so they

were the only group on which he could impose direct taxes: the existence of the Jewish commu-

nity gave the king more financial autonomy and enhanced his bargaining power with the barons.

Movement of Jews in the twelfth century seems to have been largely unregulated, but the thir-

teenth century saw more extensive regulation: each Jewish family was obliged to live in one of

about 30 designated towns containing an archa (or chest) where all contracts between Jews and

3

The intergroup contact literature began with Allport (1954); see Dovidio, Glick, and Rudman (2005) and Pettigrew and

Tropp (2012) for an overview of this literature.

1132

|

FIELDING