Over the last 100 years all cities have been buffeted by the winds of economic change. Globalisation and technological and transport developments have meant that cities have had to continually adapt, both to continue to provide jobs and contribute to national economic growth. As more traditional industries have declined, the challenge for cities has been to encourage jobs growth in new areas of the economy.

Much of the work by Centre for Cities shows the varying performance of cities in the UK. But while this is seen across a range of economic indicators, there is less collective understanding as to how this variation came about and for how long it has persisted.

Using historical data on cities this paper shows that some cities have been more successful than others at rising to this challenge. It looks at over 100 years of change in the urban areas of England and Wales, comparing cities in 1911 to their overall size and industrial make-up today.

Firstly, it sets out what has happened over the last century. It then explores why these patterns have occurred. Thirdly, it looks at whether these patterns are inevitable. And finally it offers three principles that should guide policy that attempts to support city growth over the next 100 years.

Box 1: Methodology

Defining a city

A fixed geographical definition of cities is used throughout the report based on their present day boundaries using the Primary Urban Area (PUA) definition. This is because we are interested in understanding how history affects the modern-day performance of a city, starting in 2013 and working back, rather than looking at how a city has expanded through time. 1911 administrative boundaries were matched onto current PUA boundaries. In instances where 1911 authorities straddled current PUA boundaries, only those authorities that had more than 50 per cent of their area in the PUA were included.

In total, 57 cities across England and Wales are looked at in this research. Milton Keynes and Telford are excluded because they were not established until the 1960s, and Scottish cities and Belfast are not studied because of a lack of data availability.

More detail on the PUA definition can be seen at here.

Collating jobs data for 1911

Data on occupations from the 1911 Census is used to look at the industrial structure of cities a century ago. The Census did not report occupational breakdowns for urban authorities below 5,000 in population or those districts classed as rural districts. Despite their classification, some rural districts – especially those that contained collieries – contained a large number of jobs in them, and not taking account of this would inflate the jobs growth figures presented in this report.

The Census did present figures for aggregated rural districts for every county in England and Wales. This allowed an estimation of the number of jobs and industrial structure of individual rural districts by doing the following:

  • Firstly, the total population of each missing district (for which data is available) was multiplied by the average share of people in work for the aggregated rural districts of each county. This gave an estimate of the total number of jobs.
  • Secondly, the estimated total number of jobs was multiplied by the share of jobs taken by women for the rural districts of each county. The total estimated jobs taken by men and women were then multiplied by the share of jobs in each industry by gender for the aggregated rural districts of each county. This gave an estimate of the industrial structure of each of the districts. Making the gender adjustment was important because of the very different occupations of men and women in 1911.

The data used is residence based, so it is assumed that those people living in the PUA also worked in the PUA.

Industry definitions

Industry definitions are listed in the Appendix.