If the Congressional Budget Office’s projections often turn out wrong, it’s fair to ask how much credibility one ought to assign to those projections. Gene Epstein tackles that topic in his latest “Economic Beat” column for Barron’s.

The wobbliness of the CBO’s forecasts was recently exemplified when it declared in its May projections that the projections released in February had overestimated the deficit for the current fiscal year by a whopping $200 billion. If the agency could perform that much of a flip-flop in so short a time about the current year’s outlook, what does that tell us about the reliability of its long-term projections?

Answer: The CBO projections over the long term should be treated not as forecasts, but as scenarios plausible enough to be taken seriously. The agency has been telling us that the budgetary ship of state could hit an iceberg, not that it inevitably will. And that warning comes from an organization that is Keynesian in its traditions. We know the CBO is Keynesian because it tends to put outsize weight on the positive effects on economic growth of a ballooning deficit and the negative effects of a shrinking deficit. So when this agency warns that accumulated deficits could lead to a ratio of debt to gross domestic product that poses dangers to the economy, attention should be paid.

When I reported in this column last week that, based on its most realistic projection (its “alternative fiscal scenario”), the CBO projected an 83% debt/GDP ratio by 2023, this meant only that the ratio could continue to climb from its current 75%. And when I reported in February that the same ratio could climb toward 200% over the following two decades, that was based on CBO projections that are keyed off the demographically driven explosion in entitlement spending on retiring baby boomers.

When the CBO updates those scary projections (probably next month), we must note that whatever it comes up with would look even worse if it did not come from Keynesian analysts. That’s because it develops its GDP forecasts without assuming that the soaring debt just might conceivably retard growth in GDP. If the agency did make that more realistic assumption, the debt/GDP ratio would climb toward 200% even faster.