# Instrumental Variables

In this third post I want go into the details of a method mentioned in my previous post about Structural Causal Modeling called Instrumental Variables (IV). IVs have been traditionally used to estimate structural econometric models but during the last decade there have been advocates encouraging the broader use of IV within social sciences (Bollen, 2012; Sovey & Green, 2011). One of its important characteristics is that IVs allow researchers to estimate causal relationships when randomized experiments are not feasible, so we need to use observational data.

**Instrument**

We can observe the setup of an IV () in figure 1. As researchers let’s say that we are interested in calculating the causal effect of on (). From our understanding of causal models to know the correct causal effect of we need to control for the confounders (in this case represented by the set of variables ) because if we don’t there will be confounder bias in our estimation. But what can we do if the set of variables is unobserved? Well, IVs allow us to estimate the causal effect of on without measuring the set of confounders as long as we can define the instrument variable . The variable is an instrument if:

- defines a causal relationship with .
- affects to the outcome variable only through its effect on (exclusion restriction).

Figure 1

To illustrate a practical example of the use IVs let’s take a look at the work done by Coviello et al. (2014). In this paper the authors studied the process of emotional contagion in a massive social network site, they wanted to know if the online emotional expression of users in Facebook has an effect in the emotions their friends share in the same platform. To simplify the example with a particular case and following the setup of figure 1, the variable is defined as the average emotional expression of Yvonne’s friends’ status in Facebook and variable is the emotional expression of Yvonne. Both and are measured in the same period of time. For the instrument a variable is needed that (1) defines a causal relationship with the emotional expression of Yvonne’s friends, and (2) follows the exclusion restriction (affects Yvonne’s emotional expression through the emotions of her friends). The researchers picked the rainfall as instrument , their justification was that if there is a relationship between rainfall and emotion expression it must be because emotions can’t affect the rainfall, this rationale complies with (1). For the exclusion restriction (2) they computed the model when there was no rainfall in Yvonne’s city so the instrument had no relationship with her emotional expression. The final general result showed that rainfall directly influences the emotional content status messages of users on Facebook, and it also affects the status messages of friends in cities that are not experiencing rainfall.

**Analysis**

The analysis of causal effects using IVs uses the additive properties of SEM to calculate the causal effect when we can define an instrument.

Figure 2

We already now that figure 2 represents a setup for IVs. Here, we are interested in the causal effect of on . The variable is an well-defined instrument, the set represents the group of unobserved confounding variables, and a, b, c, d the causal effects of the DAG assuming linear equations (if you need to refresh concepts like confounders or DAG take a look at my previous post).

Because and are unconfouded the causal effect of on can be estimated from the regression

(1)

Likewise, the variables and are unconfounded because the path is blocked in the collider . So the estimate of the regression of on will be equal to the direct effect on the path which is the product of the path coefficients .

Since we estimated the values and . The causal effect from to can be obtained by dividing the previous estimations (Pearl & Mackenzie, 2018)

**Limitations**

Even though the use of IV might seem a good approach to fix problems as confounding issues, the truth is that instruments can be very difficult to find. Researchers cannot use the actual data to find IVs, the knowledge about theory and the model derived from it are key to define a valid IV for the problem under investigation. A well-defined IV should not be affected by other variables in the model, this can’t be proved so it is based on the knowledge of the system studied. The correlation between the IV and must be significant because weal correlations can lead to misleading estimates for parameters and standard errors.

**An Instrumental Variable Example**

I present an example of Instrumental Variables with simulated data to show how to implement the previous analysis.

The process has the following steps:

- Plot the IV model.
- Data simulation of the DAG defined.
- Check correlation between instrument and variable .
- Estimate causal effect of IV model.
- Show the mistakes of estimation if we don’t use IV for this example.

**References**

Bollen, K. A. (2012). Instrumental Variables in Sociology and the Social Sciences. *Annual Review of Sociology*, *38*(1), 37–72. https://doi.org/10.1146/annurev-soc-081309-150141

Coviello, L., Sohn, Y., Kramer, A. D., Marlow, C., Franceschetti, M., Christakis, N. A., & Fowler, J. H. (2014). Detecting emotional contagion in massive social networks. *PloS One*, *9*(3), e90315.

Pearl, J., & Mackenzie, D. (2018). *The Book of Why: The New Science of Cause and Effect*. Basic Books.

Sovey, A. J., & Green, D. P. (2011). Instrumental Variables Estimation in Political Science: A Readers’ Guide. *American Journal of Political Science*, *55*(1), 188–200. https://doi.org/10.1111/j.1540-5907.2010.00477.x