LSE research has made a substantial contribution to the accuracy of information available to Scottish voters, ensuring transparency in any future democratic process.
What was the problem?
If Scotland became an independent country (as it nearly did in 2014), a Scottish state would need to be established.
In late May 2014, the Chief Secretary to the UK Treasury (Danny Alexander) published an estimate of the transition costs for creating a Scottish state as £2.7 billion. The number was produced by multiplying 180 agencies relevant for Scotland by the average cost for a Whitehall department reorganisation (£15 million). This figure was taken from detailed work carried out in 2010 by Professor Dunleavy and Professor White entitled “Making and breaking Whitehall departments: a guide to machinery of government changes”.
In this report, White and Dunleavy developed a methodology for tracking different cost components of reorganisations, interviewed 35 very senior civil servants, and demonstrated that reorganisation had been ill-planned and costly, perhaps especially under the Blair/Brown governments. This report was influential at the time and subsequently has been well-cited.
The Treasury had in fact made a mistake and had mis-interpreted the report. Whitehall departments are substantial bodies with hundreds or thousands of staff, whereas many of the bodies included on the list of 180 agencies had tiny staffs relevant to Scotland, for example, the Electoral Commission has six people in Scotland, who are already at work there.
Dunleavy calculated that the Treasury number of £2.7 billion was over-scaled by about 12 times, and that the immediate institutional costs of Scotland’s transition to independence would probably be around £200 million.
What did we do?
With modest funding from the Scottish Daily Press, LSE Public Policy Group with Professor Dunleavy undertook a more detailed study of transition costs, also considering the “disentangling costs” around gradually moving Scottish services out of UK administration (for example, tax and social welfare) and Scotland’s exposure down the line to “modernisation costs” where major UK services were in need of updating. This validated the £200 million immediate number, but with disentangling came to somewhat more, and in total over 10 years might amount to somewhere under £1 billion. The research team also carefully considered all the other cost projections available to Scottish citizens voting in the referendum (including work internationally) and tried to show that mostly they had no basis in any close analysis.
During this work Dunleavy visited Scotland and had a long discussion with First Minister Alex Salmond and his Deputy (now First Minister) Nicola Sturgeon and their staffs. Dunleavy learned that the First Minister and his Deputy had done no prior research on the issue, and that all UK civil servants and institutions in Whitehall and the Bank of England had refused to talk to them as UK government policy was that independence should not take place.
Dunleavy supplied an advance copy of the full report to the Scottish government ahead of publication.
What happened?
In autumn 2016, the SNP party (not the government) set up the Stewart Commission to examine the overall economic costs (or benefits) associated with Scotland becoming independent, including coverage of the government transition costs. They commissioned a paper from LSE Public Policy Group which was completed in spring 2018 and sent to the Commission. In May 2018, Professor Dunleavy’s analysis was fully incorporated in the final Stewart Commission report from the SNP.
To avoid previous press misrepresentations, simplified definitions were used. “Transition costs” is now “first five-year costs”, starting two years before independence and stretching three years afterwards. After this “long-term disentangling costs” and “long-term modernisation costs” are listed (but not costed).
The researchers derived detailed comparative costing from other similar-sized countries (Denmark, Sweden, Finland, Singapore and New Zealand). Every function was transferred, and derived costs estimated from the UK-specific data. It is estimated that the five-year transition costs (as defined above) will be no more than £450 million to £500 million.
The SNP has not yet committed to holding a second referendum. However, Dunleavy’s project ensures that voters have far more detailed and accurate information at their disposal on the likely government costs of a transition to independence.