* Got data from VP Rates as Numbers
Then ran Histogram on it to generate Figure 2.

FREQUENCIES
  VARIABLES=NFA 
  /FORMAT=NOTABLE
  /NTILES=  4
  /HISTOGRAM=NORMAL
  /STATISTICS=STDDEV MINIMUM MAXIMUM MEAN MEDIAN
	SKEWNESS KURTOSIS
  /ORDER=  ANALYSIS .

graph
    /scatterplot(bivar)  VC with NFA.
/missing=LISTWISE.

* Figure 4--scatterplot part 2.
* did this after filtering out middle counties.

graph
    /scatterplot(bivar)  VC with NFA by type.
/missing=LISTWISE.

* regression for each part.
* filtered out just low, then just high.

REGRESSION
  /MISSING LISTWISE
  /STATISTICS COEFF OUTS R ANOVA
  /CRITERIA=PIN(.05) POUT(.10)
  /NOORIGIN 
  /DEPENDENT NFA
  /METHOD=ENTER VC.