MORGAN STANLEY: Here's How Much The Economy Will Suffer Every Week Of A Government Shutdown

There are two really interesting nuggets.

Frist are the odds of a shutdown, which they put at 25%:

Our intuition is that the probability of a shutdown is relatively low but not negligible, say one-in-four. There is a powerful incentive for cooperation ahead of mid-term elections next year, as the public already holds the Congress in low repute. In addition, Speaker Boehner has the leverage that comes from the knowledge that he holds a job that no one else wants, at least for now, and that the public is likely to blame his party for failure to come to terms. As a result, he does not need all of his caucus to approve the CR as long as enough Democrats are willing to vote for it.

And here's what it would mean for the economy if there is a shutdown:

In terms of macroeconomic consequences of a short shutdown, the sudden drop and subsequent recovery in activity would be absorbed in the same quarter. There is, however, a direct arithmetic impact on GDP. Compensation of nondefense employees and civilian defense employees makes up about one-fifth of real federal spending and about 1.5% of GDP. Eliminate a third of that in a shutdown as non-exempt workers stay home, and GDP is haircut 0.5%. Annualized, this reduces quarterly GDP growth by around 0.15 percentage points per week of shutdown. Even if ex post legislation makes up the missed pay and therefore avoids a hit to personal income, the real numbers will be gone for good because the hours worked will not be made up. Note that the Federal Reserve is not part of the federal budget, so its operations will continue uninterrupted. As it is the fiscal agent of the Treasury, debt issuance and repayment will continue without a hitch.

There's one other wrinkle which matters to us (and markets) a lot: If there's a government shutdown, then no jops report on Friday.