9/21/2009

What is the relationship between Government's share of health care spending and the growth in real spending?

This regression uses the log of health care expenditures per capita (using purchasing power parity in terms of dollars) on the long of the percentage of health care spending from the government, real per capita income, real per capita income squared, and year and country fixed effects. Using all the data available from 1960 to 2007 for the OECD countries, each one percent increase in a government’s share of health care expenditures increases health care expenditures by about 0.4 percent.

The entire estimate is available here.

. xi:reg lnhealthexppercap lnperchealthexppub percapitainc percapitainc2 i.year i.country
i.year _Iyear_1960-2007 (naturally coded; _Iyear_1960 omitted)
i.country _Icountry_1-30 (_Icountry_1 for cou~y==Australia omitted)

Source | SS df MS Number of obs = 969
-------------+------------------------------ F( 79, 889) = 703.36
Model | 1246.26407 79 15.7754946 Prob > F = 0.0000
Residual | 19.9391929 889 .022428788 R-squared = 0.9843
-------------+------------------------------ Adj R-squared = 0.9829
Total | 1266.20327 968 1.30806123 Root MSE = .14976

------------------------------------------------------------------------------
lnhealthex~p | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
lnpercheal~b | .4116497 .0477609 8.62 0.000 .3179124 .5053871
percapitainc | .0000428 6.03e-06 7.09 0.000 .0000309 .0000546
percapitai~2 | -5.44e-10 8.48e-11 -6.42 0.000 -7.11e-10 -3.78e-10
.
.
.
_cons | 2.413152 .1959603 12.31 0.000 2.028554 2.797751
------------------------------------------------------------------------------

Labels:

2 Comments:

Blogger Unknown said...

I don't use STATA so perhaps I'm not reading this correctly, but doesn't that say per capita spending has a P-value of .31?

9/24/2009 3:01 AM  
Blogger John Lott said...

No, it is the 95% confidence interval. Look at the headings.

9/24/2009 10:45 AM  

Post a Comment

<< Home