Heavy Lifting - thoughts and web finds by an economist
     I also contribute to Division of Labour Load HL's Front Page
Friday, June 01, 2007

A little more on the Peace Index

While gathering some data for another project I am working on I came across the domestic interest rates as reported in the International Financial Statistics series of the IMF. I was able to collect the interest rate data and match it up with the peace index. Unfortunately, there are countries in both series that are not included in the other.

However, in the end I was able to match 80 countries with the 2007 peace index value and their 2006 domestic lending rate. One simple hypothesis would be that interest rates should be lower in more peaceful countries because uncertainty is lower, ceteris paribus. Therefore, if the peace index is measuring what it claims to measure, we should see a positive correlation between the peace index value (higher numbers are less peaceful countries) and interest rates.

Here's a scatter plot of the data:



There is a slight upward trend in the data (I note that I have dropped the extreme outlier of Zimbabwe which had an interest rate of 486%, the only country with lending rates over 100%).

Here is the correlation between the two variables. The significance of the relationship is a bit weak (p-value of 0.10), but it is nevertheless there.
. pwcorr intrate06 peace07,sig

| intra~06 peace07
-------------+------------------
intrate06 | 1.0000
|
|
peace07 | 0.1844 1.0000
| 0.1015
|

For giggles, here's a simple regression model which relates the 2006 interest rate to the 2007 Peace Index and the 2005 interest rate:

reg intrate06 peace07 intrate05

Source | SS df MS Number of obs = 80
-------------+------------------------------ F( 2, 77) = 477.61
Model | 218250.524 2 109125.262 Prob > F = 0.0000
Residual | 17593.282 77 228.484182 R-squared = 0.9254
-------------+------------------------------ Adj R-squared = 0.9235
Total | 235843.806 79 2985.36464 Root MSE = 15.116

------------------------------------------------------------------------------
intrate06 | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
peace07 | -9.857491 4.95846 -1.99 0.050 -19.73105 .0160654
intrate05 | 1.983625 .0653942 30.33 0.000 1.853409 2.113842
_cons | 6.457444 9.966791 0.65 0.519 -13.38897 26.30386
In this case, we find that the interest rates are negatively related to the peace index which doesn't make much sense. This makes me wonder what the peace index is measuring and whether the market views "peace" the same as the "peace index."

If we include as many lagged values of the interest rate as possible (given the data I gathered), the relationship between the peace index and domestic lending rates is still negative only now insignificant:

. reg intrate06 peace07 intrate05 intrate04 intrate03 intrate02 intrate01 intrate00 intrate
> 99 intrate98 intrate97

Source | SS df MS Number of obs = 75
-------------+------------------------------ F( 10, 64) = 1221.77
Model | 234378.874 10 23437.8874 Prob > F = 0.0000
Residual | 1227.7516 64 19.1836187 R-squared = 0.9948
-------------+------------------------------ Adj R-squared = 0.9940
Total | 235606.626 74 3183.87332 Root MSE = 4.3799

------------------------------------------------------------------------------
intrate06 | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
peace07 | -1.375817 1.645092 -0.84 0.406 -4.662266 1.910632
intrate05 | 1.385047 .3440727 4.03 0.000 .697683 2.072411
intrate04 | 1.170315 .2905955 4.03 0.000 .5897834 1.750846
intrate03 | -1.315997 .1681404 -7.83 0.000 -1.651896 -.9800976
intrate02 | .5755923 .0920535 6.25 0.000 .3916942 .7594904
intrate01 | -.1926117 .2229543 -0.86 0.391 -.6380141 .2527908
intrate00 | -.9988032 .2127621 -4.69 0.000 -1.423844 -.573762
intrate99 | .3458499 .1944291 1.78 0.080 -.0425668 .7342666
intrate98 | .1010942 .1064238 0.95 0.346 -.1115118 .3137001
intrate97 | -.0014019 .0218359 -0.06 0.949 -.0450241 .0422204
_cons | 3.832371 3.087061 1.24 0.219 -2.334742 9.999484


I wonder if the problem is that many of the elements that comprise the peace index involve foreign (and sometimes distant) intervention and which could naturally mitigate the relationship between the peace index and interest rates. The intuition is that much of what makes the United States, for example, less peaceful (according to the Peace Index) happens thousands of miles away from the United States where the domestic lending rates are determined. In the absence of additional terrorism on the U.S. mainland, the actions in Afghanistan and Iraq, which influence the peace index, may have little to do with local lending rates in Iowa.

Stata Data File

Comments:
How about Index of Economic Freedom versus the Peace Index?
 
That's my next step - perhaps later tonight or tomorrow.
 
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

- A little more on the Peace Index




Archives


home


Visitors
Site Meter Blogroll Me!



Credits

Modified maystar design
powered by blogger