top of page
Writer's picture Filippos Papasavvas

How does one measure interregional inequality, and where does the UK rank?

There are many different ways of measuring interregional inequality, which can make it hard to make cross-country comparisons. The McCann (2019) paper compares the degree of the UK’s interregional inequality to that found in France, Germany, Italy, South Korea, Sweden, Japan and the USA, using different methodologies and datasets. Subject to the method used, the UK ranks from first to fourth most unequal among those eight countries.



There are many imperfect ways of measuring interregional inequality…

McCann (2019) distinguishes between two data types that can be used to measure a country’s interregional inequality: GDP per capita (or its related index, GVA per capita) and regional disposable income (RDI) per capita. Both have shortcomings:

  • GDP per capita calculates the amount of value created at the workplace, instead of the residence location. Consequently, if people from one region travel to another for work, the value they create will be attributed to the latter. This can result in false conclusions that the residents of the emigrating region are much poorer than the ones in the immigrating one. Importantly, such migration patterns are not atypical if we are to think of large city centers such as London, where many people commute for work from different parts of the country.

  • RDI per capita is derived from people’s wage and salary income, and it, therefore, excludes wealth inequality. Moreover, it can be easily influenced by fiscal policy which redistributes wealth across regions. In this case, RDI measures can fail to capture the economic dynamism of different regions, as they reflect instead government policy.

Even though the author considers GDP per capita to be an overall superior indicator of a region’s comparative wealth, this is unlikely to be true for all cases.


…and there are many ways to split a country

The way one splits a country into regions can have large implications on subsequent regional comparisons and calculations. At the same time, data limitations do narrow down significantly one’s options. For GDP and RDI per capita interregional data there are mainly three types of data sources, all provided by the OECD. These are the Territorial Level 2 classification (TL2), Territorial Level 3 classification (TL3), and Metropolitan Urban data. For this piece, I will focus on TL2 and TL3, where TL2 offers the more granular segregation.


… and there are also many ways to use GDP, RDI, and regional data!

McCann (2019) separates between five methods of using the GDP and RDI per capita data to measure a country’s interregional inequality. These are presented below with the example of GDP. The exact same method can be used with RDI data:

  1. The ratio of the highest and lowest GDP per capita region (M1)

  2. The difference of the highest and lowest GDP per capita region, divide them by the country’s average GDP per capita (M2)

  3. The ratio of the highest 10% and lowest 10% GDP per capita regions (M3)

  4. The ratio of the highest 20% and lowest 20% GDP per capita regions (M4)

  5. The interregional Gini coefficient (M5)

Where does the UK rank?

The paper compares the UK’s interregional inequality with that of France, Germany, Italy, South Korea, Sweden, Japan, and the USA, using all discussed methods, data types, and regional segregations. The outcomes of this exercise are presented in Figure 1 below, where the UK was found to always rank the most unequal at the TL3, but not at the TL2 level. At the TL2 level, the UK was occasionally surpassed by Italy, Spain and the US. Data imperfections aside, the presented analysis does indicate that the UK is one of the countries with the highest interregional inequality within the group. Outcome variability also highlights how one may reach different conclusions, depending on the data and method used.


Figure 1: The UK always ranks as the most unequal country at the TL3 level, but not at the TL2 level

Data: Bonsai Economics, McCann (2019). Note*: no USA data available. Note**: no France, Germany, Italy, Spain, USA, data available.

Comments


bottom of page