August 05, 2014
The Post’s Wonkblog has a piece telling us that we should thank the recession for the slowdown in health care cost growth. I was one of those in the camp who thought the recession was responsible for the slowdown in health care growth in 2008-2010, however I think the explanation weakens as time goes on and costs continue to grow slowly.
The point is simple. Suppose that you have $10k slashed from your income in 2008 compared to its 2007 level. We might expect that you would spend less on health care and everything else in 2008. Suppose that your income in 2009 is again $10k below where you expected it would have been back in 2007. This happens again in 2010, 2011, 2012, 2013, and 2014. In other words, your income grows at more or less the same pace that you would have expected in each of these years, but the level in each year is 10k below what you had expected it would be in back in 2007.
In this story, which more or less captures the recovery, we might expect that the level of health care spending in these later years would be lower than had been projected in 2007, but the growth rate would be pretty much the same. The Post piece tells us that ain’t so.
It cites two studies. Since one is behind a paywall, I will focus on the Brookings study which is freely available to the unwashed masses. This study finds a reasonably strong link between health care spending and GDP growth, however there is a long lag. The regressions for the growth of per capita health care spending use as independent variables current GDP growth and 5 lagged GDP terms using annual data. What is striking is that the strongest effect shows up on the fourth lagged term.
This is noteworthy in the current context because in 2013, the fourth lagged term gave us 2009 GDP growth, which was -2.8 percent. The fourth lagged term this year would give us 2010 GDP growth, which was 2.5 percent. The difference between these two implies a predicted rate of health care cost growth that is 1.6 percentage points higher in 2014 than in 2013. (This calculation uses the coefficients from column 1 of Table 1, the uptick in predicted cost growth would apply for all the regressions whose results are shown in the table, although the size would vary.)
The point is that if this study is the basis for expecting a sharp slowing of health care costs due to the weak economy, the period during which that would be true is over. Based on the study’s findings we should be seeing substantially more rapid increases in health care costs in 2014 than we did last year. Thus far this doesn’t appear to be the case, which may cause us to question the usefulness of this model for explaining recent patterns in health care cost growth.
Addendum:
Medgeek was good enough to send me the other study, a paper by David Dranove, Craig Garthwaite, and Christopher Ody, which I quickly read through. Looks to me like it provides good evidence that the recession was the major factor in reducing cost growth in 2008-2010. Their model shows that the recession would not lead to any further decline in cost growth in 2011 or later years (see Exhibit 3). In fact, the modest uptick in the employment to population ratio in subsequent years means that we should have been seeing somewhat above trend increases in health care costs in 2012-2014. So yes, there is good reason to believe that the recession was the major factor behind slower health care costs in the years 2008-2010. The continued slow growth over the last three and a half years requires another explanation.
Comments