Scottish home prices: compatible with Euro membership?

• Author: Miles, William

1. Introduction

   Monetary policy is easier to implement, and housing cycles less volatile, when different regional housing markets in a monetary union exhibit a high degree of co-movement in their fluctuations. In the UK, if home values in Wales are rising towards bubble levels while those of Scotland are contracting, the Bank of England's policy will not be optimal for both regions. The loose policy that could help Scottish prices recover would exacerbate price increases in Wales, leading to financial vulnerability and resource misallocation in the former. In contrast, the restrictive policy that would be best for Wales could send Scottish prices lower at the worst possible time for homeowners and their creditors.

   While the BOE doesn't formally target home values, the housing market has relevance for policy through its impact on the business cycle of the real economy.

   House prices can affect the real economy through wealth effects, as well as through an impact on credit availability and spending via the financial accelerator. The global financial crisis of 2007-09 was sparked by a downturn in United States housing (indeed, for the US, "Housing IS the business cycle", according to Learner (2007, 2015)). Due to housing's importance to the economy, and the division of the UK into thirteen different regions, there have been numerous studies of house price co-movement in the United Kingdom (see Ashworth and Parker (1997), Cook (2003), Meen (1999) and MacDonald and Taylor (1993) for just several of many examples). Some studies find Scotland has a relatively low level of co-movement with the rest of the UK (Drake (1995), Abbott and DeVita (2013)).

   A possibility for Scotland is that of eventually joining the euro zone. There was a referendum in 2014 on Scottish independence, which was defeated. However, the passage of Brexit in 2016 and the December 2019 parliamentary elections in which the Prime Minister Boris Johnson won a large majority and promised to complete Brexit and leave the European Union were followed by calls from Nicola Sturgeon, leader of the Scottish National Party, for another referendum on independence. Thus Scottish independence, accompanied by an attempt to stay in the EU, is a continued possibility. Should Scotland obtain independence and join (or re-join, as it were) the EU, it would be obliged to eventually adopt the euro currency. All EU members are officially obliged to do so sooner or later, with exceptions made only for the UK and Denmark.

   This brings up the issue of how Scottish house prices would co-move with other euro housing markets. Would Scottish house prices display a high level of cyclical synchronization with euro country house prices, which would help make Scotland part of an "optimal currency area" in the euro zone? In contrast, if Scottish house prices did not co-move much with those of other euro countries, the region could find monetary policy from the ECB, which would be formulated with little regard for Scotland's conditions, highly problematic.

   We accordingly analyze how closely Scottish house prices co-move with those in other euro zone nations. First, we examine correlations between Scottish and euro country house prices over our whole sample, and then calculate rolling correlations to see how linear co-movement has evolved over time. We then calculate the same measures for Scottish home values and those of the UK for comparison. As correlation is a purely linear measure, we next apply metrics developed by Mink, Jacobs and DeHaan (2012), called synchronicity and similarity. The synchronicity measure indicates whether two variables are in the same, or a different cyclical phase. That is, are both in contraction or expansion, or is one growing while the other is contracting? The synchronicity measure does not account for differences in cyclical amplitudes, however. For instance, if one region is in a mild recession, but another is in a much more severe downturn, their cycles are different, although the synchronicity measure would suggest a high level of co-movement. Moreover, optimal policies for each region would be different; the first region would benefit from a mild monetary easing, while the second would be best off with much stronger counter-cyclical measures.

   Accordingly, Mink, et al. (2012) have developed a second measure called similarity, which takes account of differences in the amplitude of cycles between regions. Two regions could be in the same cyclical phase, but if their cycles differ substantially in magnitudes, they will score low in terms of co-movement by the similarity metric.

   To anticipate our results, we find that while Scottish home prices have a greater linear association with those of the UK than those of most large economies in the euro zone, the difference is not that large for some euro countries, such as Finland and France. Moreover, using the synchronicity measure, Scottish home values appear to co-move just as much, overall, with their euro counterparts as they do with those of the UK, depending on the sample. All of this would suggest Scottish home prices would lose little, in terms of co-movement with other regions in a central bank jurisdiction, by leaving the UK and joining the euro.

   However, when we employ the similarity measure, Scotland exhibits much greater divergence from the collection of euro economies than previous measures, which failed to take account of different amplitudes, indicated. Scotland is next-to-last among the nine regions, exhibiting more co-movement than only Spain, which of course had a well-known housing boom and bust episode, and even less than Ireland, which suffered similar difficulties. Overall, the results indicate that for the Scottish housing market, leaving the pound could entail substantial risk.

   This paper proceeds as follows. The next section describes the previous literature. The third details the data and methodology. The results are described in the fourth section, and the fifth concludes.

2. Previous Literature

   There have been numerous studies on house price co-movement between different regions in the UK, and a smaller but substantial literature on such co-movement in the euro zone. The UK consists of thirteen separate regions, one of which is Scotland, with house price indices for all compiled by the Nationwide Building Society. One part of the literature on UK regional house price co-movement has focused on what is termed the "ripple effect". The issue with the ripple effect is whether house price shocks in London or the South East of the UK create "ripples" in the home values of other regions (see MacDonald and Taylor (1993), Drake (1995), Ashworth and Parker (1997), Meen (1999) and Cook and Thomas (2003) for some examples).

   Some papers examine other types of house price interaction, such as...