Heavy Lifting - thoughts and web finds by an economist
     I also contribute to Division of Labour Load HL's Front Page
Friday, July 10, 2009

NEA grants as political patronage?

This report provides data on the amount of grant money doled out by the NEA for 2009.

I grabbed the data and mixed with the percentage of the vote received by President Obama at the state level. The first take is looking at total state level grants related to a dummy variable that takes a value of one if the state was 50% or more for Obama:


. reg totstate obamab if firstobs&!missing(obama),r

Linear regression Number of obs = 51
F( 1, 49) = 5.07
Prob > F = 0.0289
R-squared = 0.0851
Root MSE = 9.7e+05

------------------------------------------------------------------------------
| Robust
totstate | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
obamab | 581481.5 258233.4 2.25 0.029 62542.04 1100421
_cons | 275000 58516.78 4.70 0.000 157406.1 392593.9


The simple difference in means test suggests that states that went for Obama received about $600,000 more in NEA grants than states that did not go for Obama. States that did not vote for Obama get only $275,000 on average.

Okay - but maybe the states that went for Obama are larger and therefore they applied for and received more grants, thereby increasing their average.

Here is a simple difference in means test looking at the number of grants per state:


. reg numgrants obamab if firstobs&!missing(obama),r

Linear regression Number of obs = 51
F( 1, 49) = 5.11
Prob > F = 0.0283
R-squared = 0.0852
Root MSE = 22.524

------------------------------------------------------------------------------
| Robust
numgrants | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
obamab | 13.49537 5.972905 2.26 0.028 1.492368 25.49837
_cons | 5.208333 .9332125 5.58 0.000 3.332973 7.083694
------------------------------------------------------------------------------


States that did not go for Obama received approximately 5 grants and those that went for Obama got approximately 18 (13+5) grants. These grants are the outcome of applications and decisions - so the reduced form estimates here do not answer the question of whether the grants are patronage or not. It might be the case that the states that didn't vote for Obama also don't apply for NEA grants, and those states that did vote for Obama have more culture in general and apply for more grants. There are many on the left who would argue that this is the case

Unfortunately, we don't have data on the number of grants applied for - that is where the real test of this hypothesis could be obtained.

Nevertheless, we can look at potential form of patronage - average dollars per grant awarded. If the hypothesis is true then one would expect that the average size of the grants awarded to Obama voting states would be larger than in those states that did not vote for Obama.

Here is a difference in means test looking at the amount of the average grant:


. reg avegrant obamab if firstobs&!missing(obama),r

Linear regression Number of obs = 51
F( 1, 49) = 0.03
Prob > F = 0.8581
R-squared = 0.0007
Root MSE = 15244

------------------------------------------------------------------------------
| Robust
avegrant | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
obamab | -768.6525 4276.216 -0.18 0.858 -9362.03 7824.725
_cons | 48762.31 3109.118 15.68 0.000 42514.3 55010.31

Here we see that there is no statistically significant difference in the average grant size awarded in Obama-voting states and that in non-Obama-voting states. This might be evidence against the grants-as-patronage but is not guaranteed. It might be that the average grant size is not where the patronage manifests itself.

I did this in the 10 minutes or less time limit I impose upon myself for these little exercises. It will be an interesting masters thesis (or dissertation) that looks deeper into these patterns to determine if political patronage explains the variation across states - as has been shown to be the case with Lincoln, FDR, and other presidents.

Stata v9 data set

Labels: , ,

Comments: Post a Comment



Purchase




Le Chai - galerie du vin



Support



Popularity

Posts that contain Craig Depken per day for the last 90 days.



Contacts

Heavy Lifting's Main Page
Email Me
Atom Feed

WWW
Heavy Lifting




Great Links

Money I Found Today

Heavy Lifting - Firehose style (56k warning)



Recent Posts

- NEA grants as political patronage?




Archives


home


Visitors
Site Meter Blogroll Me!



Credits

Modified maystar design
powered by blogger