A National Audit Office review of government reports on the likely economic of Brexit during the EU referendum campaign has found some of the Treasury’s key assumptions were wrong.
In a study of the finance department’s analysis of the impact of an EU exit, auditors highlighted that estimating the economic impact of leaving was complex as there was no precedent, and the extent of the impact depended on future government policy and international negotiations which are yet to take place.
RELATED CONTENT
In its analysis of the impact of leaving the EU, the Treasury warned that leaving the bloc could hit the public finances by £36bn a year.
Among its claims, the Treasury said the hit to British gross domestic product could range from 3.4% to 9.5% after 15 years, depending on the model the UK opted for after Brexit, and each percentage point reduction in long-term GDP there would be a reduction in tax receipts of about £7bn. The Treasury's central forecast – based on Britain negotiating a Canada-style bilateral agreement with Europe – is for GDP to be 6% lower, with tax receipts to take a £36bn-a-year hit.
The analysis was controversial in the campaign and was branded part of then chancellor George Osborne's “project fear” approach to the campaign. Senior pro-Brexit figures including, Bernard Jenkin, the then chair of the Public Administration and Constitutional Affairs Committee, has said civil servants should not have produced the document.
The NAO concluded that although the Treasury said it was not producing a forecast, the ministry made a number of assumptions in its analysis. It firstly intentionally assumed no policy response to Brexit, in order to avoid prejudging future government policy decisions, and also excluded any changes from the Bank of England.
However, since the outcome of the referendum in June 2016, the central bank announced a package of measures intended to boost the economy including an interest rate cut from 0.5% to 0.25%, and the expansion of the quantitative easing programme to buy government bonds by £60bn.
Some other key assumptions have not held in practice, the report said, including that the government would trigger Article 50 immediately after the vote when it was not triggered until March 2017.
In the report, auditors set out a series of principles for such reports from Whitehall.
They said that such interventions required an effective quality assurance process to guard against technical errors and ensure the modelling approach and assumptions were appropriate.
They added that departments should establish and maintain quality assurance processes that provide effective challenge, informing and improving the quality of estimates, as well as incorporating expert input from within government. The NAO cautioned that interventions must also be sufficiently independent, or include elements suitably independent, of the analytical processes.
The Treasury told the NAO it had followed an enhanced quality assurance approach for the report, compared with its usual approach to published analysis. This included engaging and utilising a larger number of analysts across the Treasury, who were not directly involved in the underlying analysis as well as higher level of engagement with senior officials through the senior management boards.
The department also pointed to external review by Professor Sir Charles Bean of the National Institute of Economic and Social Research, the former deputy governor of the Bank of England.