Eight quick responses to the Treasury’s short-term Brexit impact study

The Treasury today published its estimate of the short-term impact of a vote to leave the EU. Ten quick responses.

  1. The “shock” scenario assumes the central case of the Treasury’s long-term impact assessment. If that scenario is way too high (HINT: yes) that will make the short-term impact over-stated too.
  2. The “recession” claim consists of four quarters of -0.1% growth. If that went -0.2%, 0, -0.2%, 0 instead, there’d be no recession. Tenuous.
  3. I think there’d be 2-3% loss in short-term. OECD said 3% with overly pessimistic assumptions. HMT’s 3.6% is a bit high (mainly because its estimate of the 2030 loss is too high, meaning its “transition” impact is too high) but 3.6% is not ridiculous.
  4. The Treasury assumes shocks would begin from 2016Q3. There could be some impact as quickly as that, but I’d guess most of the short-term losses would be the year before & year after Brexit, which would actually happen in 2020.
  5. Earlier this year the Chancellor warned of the dangeous cocktail of risks, facing the economy, including factors such as China and oil price falls. The economy is mainly slowing because of other factors. Indeed, without the economy being slow anyway, the impacts the Treasury report estimates would get us nowhere near a recession. The talk of Brexit “causing” a recession is clearly scapegoating and the Chancellor getting his excuses in early for a downturn mainly caused by completely different things.
  6. Today’s Treasury analysis attributes much of the recent slowdown & weakness in sterling to concerns about the Brexit referendum. But the global slowdown & UK exposure thereto via our large exposure to global trade is surely the main driver of that, not Brexit.
  7. The Treasury says an instant triggering of Article 50 post-referendum would be a driver of uncertainty. In which case, maybe don’t do that?
  8. Interestingly, the HMT analysis assumes that fiscal policy would relax to accommodate the shock, but monetary policy would be unaffected.

UPDATE: I’ve worked a couple of things out that might be of interest. Despite all the talk of “a recession as bad as the early 1990s” or “GDP falling by 3.6%”, the Treasury forecasts no such thing. What it actually forecasts is GDP falling on a quarterly basis by 0.4% over four quarters. But here’s the fun bit. Because those four quarters are from 2016Q3 to 2017Q2, when we measure annual GDP we include the growth in the first two quarters of 2016 and the last two of 2017 (which the Treasury helpfully tells us). Assuming some sensible guess for 2016Q2 of around 0.4% or 0.5%, we get annual growth (yes, growth, not contraction) in 2016 of about 1.5% and 2017 is unchanged.

I wonder if anyone really explained to Osborne that his “year-long DIY recession” would mean annual GDP going up, not down?