Q. What's wrong with the following chart from the front page of yesterday's Wall Street Journal?
A. False precision, assumptions are identified, and a lack of any sensitivity analysis.
Precision to four decimal places! So the spending over ten years with Obama's plan will be $45.95 trillion?
Why should we believe that figure, when in early 2008 economists' predictions of government spending for 2008 were completely off?
This is not merely a style point. This false precision that is rarely questioned by the media presents the world as if we are making decisions under a situation of certainty. If we knew for a fact that were were deciding between exactly $45.95 trillion in spending and $39.96 trillion, then we can take a comparison shopping approach, as if we were deciding between two frying pans at Wal-Mart.
It would be more helpful to show a range of uncertainty that expands each year.
Another problem is a lack of sensitivity analysis. Any business analyst at a consulting firm would get sent back to the team room by her manager if she presented this chart with the request to show the sensitivity of these projections.
Tell me first off what are your base case assumptions on inflation, interest rates, productivity growth, etc. Then show we what happens to your numbers if we adjust those assumptions up or down by 5, 10, 15%.
We don't even know from this article if the two projections use the same set of assumptions! We have to assume that they don't, since the footnote lets us know that Obama's figures as presented don't even agree with what Ryan thinks Obama's figures are.
Not to mention the fact that both projections assume that nothing else will change over the next ten years. Are you willing to believe that? That this year we'll finally resolve our differences, agree on a budget, and no one will tinker with it for a decade? That we won't have any wars, recessions, bubbles?
 |
| From Innovation Bootcamp |
A. False precision, assumptions are identified, and a lack of any sensitivity analysis.
Precision to four decimal places! So the spending over ten years with Obama's plan will be $45.95 trillion?
Why should we believe that figure, when in early 2008 economists' predictions of government spending for 2008 were completely off?
This is not merely a style point. This false precision that is rarely questioned by the media presents the world as if we are making decisions under a situation of certainty. If we knew for a fact that were were deciding between exactly $45.95 trillion in spending and $39.96 trillion, then we can take a comparison shopping approach, as if we were deciding between two frying pans at Wal-Mart.
It would be more helpful to show a range of uncertainty that expands each year.
Another problem is a lack of sensitivity analysis. Any business analyst at a consulting firm would get sent back to the team room by her manager if she presented this chart with the request to show the sensitivity of these projections.
Tell me first off what are your base case assumptions on inflation, interest rates, productivity growth, etc. Then show we what happens to your numbers if we adjust those assumptions up or down by 5, 10, 15%.
We don't even know from this article if the two projections use the same set of assumptions! We have to assume that they don't, since the footnote lets us know that Obama's figures as presented don't even agree with what Ryan thinks Obama's figures are.
Not to mention the fact that both projections assume that nothing else will change over the next ten years. Are you willing to believe that? That this year we'll finally resolve our differences, agree on a budget, and no one will tinker with it for a decade? That we won't have any wars, recessions, bubbles?
Posts
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.