I was reading this article the other day. In it, the author presents an interesting hypothesis - that the lower cost of remittances might actually be making remittance harder and hurting consumers by making the market less appealing for institutions and thus removing the incentive to provide better service. (As an aside, this is one of the reasons I'm excited about Bitcoin; it has the potential to completely disrupt this market).

Surprisingly, while analyzing the remittance inflows over time, the author completely ignores the prevailing the exchange rate at the time. This was absurd to me as I would have imagined the exchange rate as a significant driver of total remittances. If the exchange rate is favorable (USD is stronger), more money would flow into the receiving country and vice-versa. So I decided to use publicly available data to dig into this.

I downloaded the Monthly Remittance data from the World Bank and the Monthly Exchange Rate data from Quandl. The downloaded datasets are available on my GitHub repository (link below).

Here is the monthly data plotted from January, 2003 to October, 2013 -

And here are the results of a simple linear regression of the total monthly remittance inflow on the exchange rate -

```
Call:
lm(formula = df$remit ~ xr$USD.MXN)
Residuals:
Min 1Q Median 3Q Max
-820.40 -201.46 9.34 232.80 782.26
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 2173.46 300.28 7.238 3.73e-11 ***
xr$USD.MXN -29.23 25.22 -1.159 0.249
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
Residual standard error: 328.9 on 128 degrees of freedom
Multiple R-squared: 0.01039, Adjusted R-squared: 0.002655
F-statistic: 1.343 on 1 and 128 DF, p-value: 0.2486
```

As you can see, and to my surprise, there is *no significant relationship* between monthly remittance inflows and the exchange rate (at least for Mexico)!

This suggests that the decision to send money home might be constrained by or dependent on factors other than the exchange rate; the cost of remittance being one of them.

The code and data for this analysis can be found on my GitHub repository.