Let’s set aside for a moment the politics of the shutdown.  Regardless of political views, we can all generally agree that the shutdown costs the economy something.  The bigger question is what that number is.  The national media have more or less universally been reporting a number calculated by IHS Global Insight, a Massachusetts-based economic forecasting firm, of $12.5 million per hour.  For you non-mathematically inclined readers, that works out to $300 million a day. The problem with that number is that it represents only the dollars that the federal government spends on goods and services each day.  It does not represent the actual economic costs of the shutdown.

The main reason for this is something called the multiplier effect.  One dollar of spending by any economic actor does not just raise GDP by that same dollar.  Instead, my spending is someone else’s income that the next person can spend.  If I buy a hot dog and a soda from the vendor down the street, I spend $3.  The hot dog vendor, in turn, uses those $3 in income to purchase buns for tomorrow’s sales.  (Yes, I know, the buns cost more than $3.  The idea is that part of the money to buy the buns is my $3.)  The grocery store uses those same $3 to pay the employee who put the buns on the shelf.  And so on. Thus, my initial $3 ends up creating an increase of more than $3 in GDP.

Obviously, the real word is more complicated.  Some of the $3 is taken out at each transaction in taxes, some of it is saved by various economic actors instead of spent, etc.  However, you can measure this effect.  The Congressional Budget Office generally calculates the multiplier for federal spending as a range between 0.5 and 2.5.  (Here is sample source,  but there are lots of CBO reports using the same numbers.) So this means that, in reality, the shutdown is costing the US economy anywhere between $150 million to $750 million a day in lost GDP.  In any event, the costs are almost certainly more than $300 million per day number that is being widely reported in the media.

Of course, this calculation doesn’t include the medium- and long-term savings we get for not borrowing nearly a third of that money to cover the deficit, but that is a different blog post entirely.