{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Instrumental variables"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## What is an instrumental variable?\n",
"\n",
"An **instrumental variable (IV)** is a variable that satisfies the following assumptions:\n",
"1. It **causes variation in the exposure/treatment** (i.e. relevance assumption - it is correlated with X, an endogenous explanatory variable) - also known as the *strong first stage* assumption\n",
"2. It is **unrelated to the outcome** (and so any affect on outcome is exclusively via the instruments effect on the exposure) - also known as the *exclusion restriction*\n",
"3. As it is unrelated to the outcome, is is therefore **unrelated to unmeasured confounders** (unmeasured differences in characteristics that affect outcomes) (i.e. **exogeneity** assumption - it is an exogneous variable) [[McClellan et al. 1994]](https://doi.org/10.1001/jama.1994.03520110039026)[[Igelström et al. 2022]](https://doi.org/10.1136/jech-2022-219267) [[source]](https://towardsdatascience.com/instrumental-variables-a-practical-explanation-1a583408a5b9) [[source]](https://medium.com/teconomics-blog/machine-learning-meets-instrumental-variables-c8eecf5cec95)\n",
"\n",
"This is illustrated below:\n",
"````{mermaid}\n",
" flowchart LR;\n",
"\n",
" iv(\"Instrumental variable\"):::white;\n",
" e(\"Exposure/treatment\"):::white;\n",
" o(\"Outcome\"):::white;\n",
" u(\"Unmeasured confounders\"):::white;\n",
"\n",
" iv --> e;\n",
" e --> o;\n",
" u -.-> e;\n",
" u -.-> o;\n",
" \n",
" classDef white fill:#FFFFFF, stroke:#FFFFFF;\n",
" classDef black fill:#FFFFFF, stroke:#000000;\n",
" classDef empty width:0px,height:0px;\n",
" classDef green fill:#DDF2D1, stroke: #FFFFFF;\n",
"````"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Due to these characteristics, instrumental variables enable us to mimic randomisation to treatment. [[McClellan et al. 1994]](https://doi.org/10.1001/jama.1994.03520110039026) In fact, **randomisation to treatment** in an RCT is an example instrumental variable (it meets the above assumptions).\n",
"\n",
"````{mermaid}\n",
" flowchart LR;\n",
"\n",
" iv(\"RCT randomisation to treatment\"):::white;\n",
" e(\"Exposure/treatment\"):::white;\n",
" o(\"Outcome\"):::white;\n",
" u(\"Unmeasured confounders\"):::white;\n",
"\n",
" iv --> e;\n",
" e --> o;\n",
" u -.-> e;\n",
" u -.-> o;\n",
" \n",
" classDef white fill:#FFFFFF, stroke:#FFFFFF;\n",
" classDef black fill:#FFFFFF, stroke:#000000;\n",
" classDef empty width:0px,height:0px;\n",
" classDef green fill:#DDF2D1, stroke: #FFFFFF;\n",
"````\n",
"\n",
"You have likely identified an instrumental variable 'if people are confused when you tell them about the instrument's relationship to the outcome'.\n",
"* No-one is confused if you say family size will reduce labour supply of women\n",
"* They will be confused if you use gender composition of first two children as instrumental variable, and find 'that mothers whose first two children were the same gender were employed outside the home less than those whose two children had a balanced sex ratio', as they don't expect gender composition to incentive work outside home - but it is related as if both the same gender, they're more likely to try again for child of another gender [[Causal Inference: The Mixtape - Scott Cunningham]](https://mixtape.scunning.com/06-regression_discontinuity)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Two-stage least squares (2SLS)\n",
"\n",
"You can use the two-stage least squares (2SLS) method to estimate the causal effect, using instrumental variables for individual-level data.\n",
"\n",
"1. **Regress exposure (${x}$) on instrumental variable (G) to produce $\\hat{x}$ (estimate of exposure independent of confounders)**\n",
"2. **Regress outcome (Y) on $\\hat{x}$**. Here, $\\hat{x}$ is replacing the actual value of the problematic predictor, ${x}$. [BSc Medical Sciences]\n",
"\n",
"We include the measured confounders in both stages.\n",
"* Included in first stage to reflect exogenous movement in treatment\n",
"* Included in second stage to avoid omitted variable bias\n",
"\n",
"The 'groups being compared differ only in likelihoods of treatment, as opposed to a division into pure treatment and control groups'. Hence, this 'method estimates an incremental or \"marginal\" effect of treatment only over the range of variation in treatment across the IV groups'. [[McClellan et al. 1994]](https://doi.org/10.1001/jama.1994.03520110039026)\n",
"\n",
"'IV analysis estimates a local average treatment effect (LATE) among 'compliers' - individudals whose exposure status is affected by the instrument. This group cannot be precisely identified, and the LATE may therefore sometimes be of limited practical or policy relevance'. [[Igelström et al. 2022]](https://doi.org/10.1136/jech-2022-219267)\n",
"\n",
"There is a variant called '**Three-Stage Least Squares (3SLS)**. This method is an extension of the 2SLS method and is used when there are more than two endogenous variables in the model. The 3SLS method is based on the idea that the endogenous variables are correlated with each other, and therefore, the coefficients of these variables need to be estimated simultaneously.' [[source]](https://fastercapital.com/content/Econometric-Techniques--Understanding-Tinbergen-s-Statistical-Innovations.html)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Assumptions of instrumental variable analysis\n",
"\n",
"(1) **Instrumental variable assumptions**(as above)\n",
"\n",
"(2) **Homogeneity assumption** - the association between the instrumental variable and the exposure is homogenous (same for everyone in the population), or the effect of the exposure on the outcome is homogenous [BSc Medical Sciences]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Limitations\n",
"\n",
"'A concern with any IV strategy is that the instrumental variable is correlated with unobserved determinants of the outcome of interest.' Can address by estimating models for held-back risk factors pre-determined at delivery date, looking for correlation with the IV. [[Card et al. 2018]](http://www.nber.org/papers/w24493)\n",
"\n",
"A **weak instrument** will produce an estimate that is biased and imprecise (and we may then have false confidence in it being unbiased). To help address this...\n",
"* Check strong first stage assumption - i.e. want statistically significant relationship when regress X on Z - suggests F-statistic of at least 11 in first stage.\n",
"* Can't test exclusion restriction assumption - just have to think through any possible ways that the IV could affect the outcome that aren't via the treatment. [[source]](https://medium.com/teconomics-blog/machine-learning-meets-instrumental-variables-c8eecf5cec95)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Non-parametric instrumental variable methods\n",
"\n",
"'**Most IV applications make use of a two-stage least squares** procedure [2SLS; e.g., Angrist et al., 1996] that requires assumptions of linearity and homogeneity (e.g., all airline customers must have the same price sensitivity). **Nonparametric IV methods from the econometrics** literature relax these assumptions [e.g., Newey and Powell, 2003]. However, these methods work by modeling the outcome as an unknown linear combination of a pre-specified set of basis functions of the treatment and other covariates (e.g. Hermite polynomials, wavelets or splines) and then modeling the conditional expectation of each of these basis functions in terms of the instruments (i.e. the number of equations is quadratic in the number of basis functions). This **requires a strong prior understanding of the data generating process** by the researcher, and the **complexity of both specification and estimation** explodes when there are more than a handful of inputs.' [[Hartford et al. 2017]](http://proceedings.mlr.press/v70/hartford17a.html)\n",
"\n",
"[Hartford et al. 2017](http://proceedings.mlr.press/v70/hartford17a.html) propose the Deep IV framework, which uses ML to perform IV analysis...\n",
"* First stage: fit conditional distribution for treatment given instruments and covariates with a deep neural net (stochastic gradient descent (SGD)) [[Hartford et al. 2017]](http://proceedings.mlr.press/v70/hartford17a.html)\n",
"* Second stage: 'target a loss function involving integration over the conditional treatment distribution from the first stage' [[Hartford et al. 2017]](http://proceedings.mlr.press/v70/hartford17a.html)... 'incorporate the fitted conditional distribution to minimize a loss function using a second deep neural net, and use an out-of-sample causal validation procedure to tune the deep network hyper-parameters' [[source]](https://medium.com/teconomics-blog/machine-learning-meets-instrumental-variables-c8eecf5cec95)\n",
"\n",
"They provide the [package on GitHub](https://github.com/jhartford/DeepIV) for this analysis using Python (broken - only supports Keras 2.0.6) (although there has been non-integrated pull request to make it work with Keras 2.3.1).\n",
"\n",
"[Xu et al. 2021](https://openreview.net/forum?id=sy4Kg_ZQmS7) provide another example of IV with deep learning, again [sharing code on GitHub](https://github.com/liyuan9988/DeepFeatureIV). This has open peer review, and so helpfully has some insightful comments around assumptions and theoretical justifications."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Examples\n",
"\n",
"### Example 1: Mendelian randomisation\n",
"\n",
"Disease association with non-genetic risk factors are often confounded - for example:\n",
"````{mermaid}\n",
" flowchart LR;\n",
"\n",
" e(\"Drinking alcohol\"):::green;\n",
" o(\"Lung cancer\"):::green;\n",
" c(\"Smoking\"):::white;\n",
"\n",
" e --> o;\n",
" c --> e; c --> o;\n",
" \n",
" classDef white fill:#FFFFFF, stroke:#FFFFFF;\n",
" classDef black fill:#FFFFFF, stroke:#000000;\n",
" classDef empty width:0px,height:0px;\n",
" classDef green fill:#DDF2D1, stroke: #FFFFFF;\n",
"````"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Genotype-phenotype associations are much less likely to be confounded. They are aetiological associations (causing or contributing to disease/condition development). \n",
"\n",
"If we know of a gene closely linked to the phenotype without direct effect on the disease, it can often be reasonably assumed that the gene is not itself associated with any confounding factors - a phenomenon called **Mendelian randomization**. Genetic variants (G) should influence the exposure (X) but should not be directly associated with confounders (U) or the outcome (Y).\n",
"\n",
"We can then identify the causal relationship between X and Y - if it doesn't cause Y, then G should be independent of Y. This is the same logic as we use for an RCT (where G would be randomisation to treatment).\n",
" \n",
"````{mermaid}\n",
" flowchart TD;\n",
"\n",
" con:::outline;\n",
" subgraph con[\"If X doesn't cause Y...\"]\n",
" G:::white;\n",
" X:::white;\n",
" Y:::white;\n",
" U:::white;\n",
" end\n",
"\n",
" G --> X;\n",
" U --> X;\n",
" U --> Y;\n",
"\n",
" uncon:::outline;\n",
" subgraph uncon[\"If X causes Y...\"]\n",
" G2(\"G\"):::white;\n",
" X2(\"X\"):::white;\n",
" Y2(\"Y\"):::white;\n",
" U2(\"U\"):::white;\n",
" end\n",
"\n",
" G2 --> X2;\n",
" U2 --> X2;\n",
" U2 --> Y2;\n",
" X2 --> Y2;\n",
"\n",
" classDef white fill:#FFFFFF, stroke:#FFFFFF;\n",
" classDef black fill:#FFFFFF, stroke:#000000;\n",
" classDef outline fill:#FFFFFF;\n",
"````\n",
"\n",
"[BSc Medical Sciences]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Example 2: Effect of intensive treatment on mortality in patients with acute myocardial infarction\n",
"\n",
"Example from [McClellan et al. 1994](https://doi.org/10.1001/jama.1994.03520110039026): Want to determine the effect of more intensive treatments (e.g. catheterisation and revascualisation) on mortality in elderly patients with acute myocardial infarction. They use distance from hospital as an 'instrumental variable to account for unobserved case-mix variation (selection bias) in observational Medicare claims data' [[McClellan et al. 1994]](https://doi.org/10.1001/jama.1994.03520110039026)\n",
"\n",
"Why is distance an instrumental variable?\n",
"* Intensive treatment only performed at certain hospitals\n",
"* Comparing treatments and outcomes between patients treated at different hospitals risks selection bias, 'because physician and patient decisions that influence treatment choice may ALSO influence choice of hospital to go to - e.g. AMI patients who appear to be better candidates for catheterisation may be disproportionately admitted to catheterisation hospitals' [[McClellan et al. 1994]](https://doi.org/10.1001/jama.1994.03520110039026)\n",
"* Distance from hospital (specifically, differential distance to hospital providing intensive treatment):\n",
" * **Affects** probability of receiving intensive treatment\n",
" * **Does not affect** patient characteristics\n",
"* Hence, distance from hospital can be used as an instrumental variable [[Igelström et al. 2022]](https://doi.org/10.1136/jech-2022-219267) - it meets the IV assumptions:\n",
" 1. Causes variation in whether receive intensive treatment\n",
" 2. Unrelated to mortality\n",
" 3. Unrelated to unmeasured confounders (like the physician and patient decisions)\n",
"\n",
"````{mermaid}\n",
" flowchart LR;\n",
"\n",
" treat(\"Intensive treatment (e.g. catheterisation)\"):::green;\n",
" mortality(\"Mortality\"):::green;\n",
" hosp(\"Hospital attended (i.e. whether it offers the treatment)\"):::white;\n",
" dist(\"Differential distance to hospital providing intenstive treatment\"):::white;\n",
" decisions(\"Physician and patient decisions\"):::white;\n",
"\n",
" dist --> hosp;\n",
" decisions --> hosp;\n",
" hosp --> treat;\n",
" treat --> mortality;\n",
" \n",
" classDef white fill:#FFFFFF, stroke:#FFFFFF;\n",
" classDef black fill:#FFFFFF, stroke:#000000;\n",
" classDef empty width:0px,height:0px;\n",
" classDef green fill:#DDF2D1, stroke: #FFFFFF;\n",
"````\n",
"\n",
"First, checked for presence of selection bias (i.e. whether hospital type affects treatment intensity - and therefore whether simple comparisons of treatments and outcomes across hospital types are valid). Then, as was present, used IV methods to estimate effect of intensive treatment on mortality. **Differential distance approximately randomises patients to different likelihoods of receiving intensive treatments**, uncorrelated with health status.\n",
"* First, compared two groups of apx. equal size - patients near to catheterisation hospital (differential distance <= 2.5 miles) and patients far from them (> 2.5 miles). This evidenced assumption that distribution of health status in AMI patients is independent of differential distance, and illustrates IV method, estimating the 'average effect of invasive treatment for all patients who are marginal from the standpoint of the near-far IVs - those who undergo catheterisation in the relatively near group and not in the relatively far group - if the two groups are balanced and if catheterisation is the only treatment that differs between the groups'\n",
"* Then used more general IV estimation netchniques with wide range of differential distance groups, and account for small remaining observable differences between differential-distance groups [[McClellan et al. 1994]](https://doi.org/10.1001/jama.1994.03520110039026)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Example 3: caesarean section\n",
"\n",
"This example is taken from It’s about time: Cesarean sections and neonatal health [[Costa-Ramón et al. 2018]](https://linkinghub.elsevier.com/retrieve/pii/S0167629617307609).\n",
"\n",
"The study aimed to estimate the causal relationship between caesarean section and newborn health: Apgar-1 and Apgar-5; Reanimation (assisted ventilation); ICU admission; Neonatal death; Umbilical cord pH.\n",
"\n",
"Several known measured confounders (e.g. age, gestational length, obstetric risk), but also omitted variable bias due to other unmeausred confounders.\n",
"\n",
"````{mermaid}\n",
" flowchart LR;\n",
"\n",
" e(\"Caesarean section\"):::green;\n",
" o(\"Neonatal health (e.g. Apgar)\"):::green;\n",
" c(\"Measured and unmeasured confounders\"):::white;\n",
"\n",
" e --> o;\n",
" c --> e; c --> o;\n",
" \n",
" classDef white fill:#FFFFFF, stroke:#FFFFFF;\n",
" classDef black fill:#FFFFFF, stroke:#000000;\n",
" classDef empty width:0px,height:0px;\n",
" classDef green fill:#DDF2D1, stroke: #FFFFFF;\n",
"````"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Time of birth can be used as an instrumental variable as:\n",
"* It is associated with treatment (unplanned C-sections more likely in early hours of night)\n",
"* It is unrelated to outcome and confounders (mothers giving birth at different times of day are observationally similar - suggesting excess number of C-sections observed are due to non-medical reasons)\n",
"\n",
"````{mermaid}\n",
" flowchart LR;\n",
"\n",
" e(\"Caesarean section\"):::green;\n",
" o(\"Neonatal health (e.g. Apgar)\"):::green;\n",
" c(\"Measured and unmeasured confounders\"):::white;\n",
" i(\"Time of day\"):::white;\n",
"\n",
" i --> e;\n",
" e --> o;\n",
" c --> e; c --> o;\n",
" \n",
" classDef white fill:#FFFFFF, stroke:#FFFFFF;\n",
" classDef black fill:#FFFFFF, stroke:#000000;\n",
" classDef empty width:0px,height:0px;\n",
" classDef green fill:#DDF2D1, stroke: #FFFFFF;\n",
"````"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Showing example of results for Apgar-5 scores.\n",
"\n",
"From **standard OLS estimation**, unplanned CS coefficient for predicting Apgar-5 (with standard error in parentheses, clustered at hospital-shift level):\n",
"* -0.219 (0.038) (regression controlling only for weekday and hospital fixed effects)\n",
"* -0.219 (0.037) (maternal controls added)\n",
"* -0.142 (0.043) (pregnancy controls added)\n",
"\n",
"**Interpretation**: Delivering C-section associated with decline in Apgar-5. However, likely biased as C-section v.s. vaginal birth cohorts not comparable.\n",
"\n",
"Hence, performed 2SLS. In **first stage**, early night coefficients for predicting unplanned C-section were:\n",
"* 0.073 (0.011)\n",
"* 0.073 (0.011)\n",
"* 0.063 (0.011)\n",
"\n",
"**Intepretation**: Births in early night are 6% more likely to be by caesarean.\n",
"\n",
"In **second stage**, unplanned CS coefficients for predicting Apgar-5:\n",
"* -0.965 (0.404)\n",
"* -0.987 (0.408)\n",
"* -0.936 (0.464)\n",
"\n",
"**Interpretation**: Increased probability of Apgar-5, significant at 5% significance level."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Python example\n",
"\n",
"This example is taken from [Causal Inference for the Brave and True](https://matheusfacure.github.io/python-causality-handbook/landing-page.html) (MIT license).\n",
"\n",
"Interested in causal effect of education on wage. Confounders like ability (which affect education and wage).\n",
"\n",
"````{mermaid}\n",
" flowchart LR;\n",
"\n",
" e(\"Education (years of schooling)\"):::green;\n",
" o(\"Wage\"):::green;\n",
" c(\"Year of birth State of birth Ability\"):::white;\n",
"\n",
" e --> o;\n",
" c --> e; c --> o;\n",
" \n",
" classDef white fill:#FFFFFF, stroke:#FFFFFF;\n",
" classDef black fill:#FFFFFF, stroke:#000000;\n",
" classDef empty width:0px,height:0px;\n",
" classDef green fill:#DDF2D1, stroke: #FFFFFF;\n",
"````"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Quarter of birth (i.e. dividing year into Q1, Q2, Q3, Q4) impacts education (kids born earlier in school year start education at older age than kids born later in school year), and does not impact wage (other than through its impact on education).\n",
"\n",
"````{mermaid}\n",
" flowchart LR;\n",
"\n",
" e(\"Education (years of schooling)\"):::green;\n",
" o(\"Wage\"):::green;\n",
" c(\"Year of birth State of birth Ability\"):::white;\n",
" i(\"Quarter of birth\"):::white;\n",
"\n",
" i --> e;\n",
" e --> o;\n",
" c --> e; c --> o;\n",
" \n",
" classDef white fill:#FFFFFF, stroke:#FFFFFF;\n",
" classDef black fill:#FFFFFF, stroke:#000000;\n",
" classDef empty width:0px,height:0px;\n",
" classDef green fill:#DDF2D1, stroke: #FFFFFF;\n",
"````"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Set-up (import packages and data)"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"# Import packages\n",
"from dataclasses import dataclass\n",
"from linearmodels.iv import IV2SLS\n",
"import matplotlib.pyplot as plt\n",
"import os\n",
"import pandas as pd\n",
"import statsmodels.formula.api as smf\n",
"\n",
"# Define file paths\n",
"@dataclass(frozen=True)\n",
"class Paths:\n",
" '''Singleton object for storing paths to data and database.'''\n",
"\n",
" data = '../data'\n",
" ak91 = 'ak91.csv'\n",
"\n",
"\n",
"paths = Paths()"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"
"
],
"text/plain": [
" log_wage years_of_schooling year_of_birth quarter_of_birth \\\n",
"0 5.790019 12.0 30.0 1.0 \n",
"1 5.952494 11.0 30.0 1.0 \n",
"2 5.315949 12.0 30.0 1.0 \n",
"3 5.595926 12.0 30.0 1.0 \n",
"4 6.068915 12.0 30.0 1.0 \n",
"\n",
" state_of_birth q1 q2 q3 q4 \n",
"0 45.0 1 0 0 0 \n",
"1 45.0 1 0 0 0 \n",
"2 45.0 1 0 0 0 \n",
"3 45.0 1 0 0 0 \n",
"4 37.0 1 0 0 0 "
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Convert quarter of birth from categorcal to dummy variable\n",
"factor_data = data.assign(\n",
" **{f'q{int(q)}': (data['quarter_of_birth'] == q).astype(int)\n",
" for q in data['quarter_of_birth'].unique()})\n",
"\n",
"factor_data.head()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Traditional OLS estimate"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
"def parse(model, exog=\"years_of_schooling\"):\n",
" param = model.params[exog]\n",
" se = model.std_errors[exog]\n",
" p_val = model.pvalues[exog]\n",
" print(f\"Parameter: {param}\")\n",
" print(f\"SE: {se}\")\n",
" print(f\"95 CI: {(-1.96*se,1.96*se) + param}\")\n",
" print(f\"P-value: {p_val}\")"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Parameter: 0.06732572817658422\n",
"SE: 0.00038839984390486796\n",
"95 CI: [0.06656446 0.06808699]\n",
"P-value: 0.0\n"
]
}
],
"source": [
"formula = \"log_wage ~ years_of_schooling + C(state_of_birth) + C(year_of_birth) + C(quarter_of_birth)\"\n",
"ols = IV2SLS.from_formula(formula, data=data).fit()\n",
"parse(ols)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Quarter of birth\n",
"\n",
"Show that quarter of birth is an instrumental variable, which requires it to:\n",
"\n",
"(1) **Cause variation in exposure** - individuals born in the last quarter of the year have slightly more time in education than those born in the beginning of the year\n",
"\n",
"We check this using a graph and regression. For simplicity, we're just using Q4 (yes/no) as our instrument.\n",
"\n",
"In the regression, we regress the instrumental variable (q4) and confounders (year + state of birth) on the exposure (years of schooling). This will statistically demonstrate that Q4 is an instrumental variable - will show us that quarter of birth affects years of schooling, whilst also control for year and state of birth. Here, on average, Q4 have 0.1 more years of education than those born in other quarters of year."
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"group_data = (data\n",
" .groupby([\"year_of_birth\", \"quarter_of_birth\"])\n",
" [[\"log_wage\", \"years_of_schooling\"]]\n",
" .mean()\n",
" .reset_index()\n",
" .assign(time_of_birth = lambda d: d[\"year_of_birth\"] + (d[\"quarter_of_birth\"])/4))\n",
"\n",
"plt.figure(figsize=(15,6))\n",
"plt.plot(group_data[\"time_of_birth\"], group_data[\"years_of_schooling\"], zorder=-1)\n",
"for q in range(1, 5):\n",
" x = group_data.query(f\"quarter_of_birth=={q}\")[\"time_of_birth\"]\n",
" y = group_data.query(f\"quarter_of_birth=={q}\")[\"years_of_schooling\"]\n",
" plt.scatter(x, y, marker=\"s\", s=200, c=f\"C{q}\")\n",
" plt.scatter(x, y, marker=f\"${q}$\", s=100, c=f\"white\")\n",
"\n",
"plt.title(\"Years of Education by Quarter of Birth (first stage)\")\n",
"plt.xlabel(\"Year of Birth\")\n",
"plt.ylabel(\"Years of Schooling\");"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"q4 parameter estimate:, 0.10085809272786678\n",
"q4 p-value:, 5.464829416613623e-15\n"
]
}
],
"source": [
"# Run regression\n",
"first_stage = smf.ols(\n",
" 'years_of_schooling ~ C(year_of_birth) + C(state_of_birth) + q4',\n",
" data=factor_data).fit()\n",
"\n",
"# Display results\n",
"print(\"q4 parameter estimate:, \", first_stage.params[\"q4\"])\n",
"print(\"q4 p-value:, \", first_stage.pvalues[\"q4\"])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"(2) **Be unrelated to the outcome** (and therefore to unmeasured confounders) - influence should all be due to effect on treatment. Here, we see seasonal pattern in outcome (earnings). We run a regression of the instrumental variable (Q4) and confounders (year + state of bith) on the outcome (wages). We see a significant result. Those born in Q4 have, on average 0.8% higher wages. P-value not as close to 0 as before, but pretty significant.\n",
"\n",
"**Note:** It's not possible to verify this condition - we can only argue in favour of it - i.e. no reason for pattern to affect income beyond through education."
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"plt.figure(figsize=(15,6))\n",
"plt.plot(group_data[\"time_of_birth\"], group_data[\"log_wage\"], zorder=-1)\n",
"for q in range(1, 5):\n",
" x = group_data.query(f\"quarter_of_birth=={q}\")[\"time_of_birth\"]\n",
" y = group_data.query(f\"quarter_of_birth=={q}\")[\"log_wage\"]\n",
" plt.scatter(x, y, marker=\"s\", s=200, c=f\"C{q}\")\n",
" plt.scatter(x, y, marker=f\"${q}$\", s=100, c=f\"white\")\n",
"\n",
"plt.title(\"Average Weekly Wage by Quarter of Birth (reduced form)\")\n",
"plt.xlabel(\"Year of Birth\")\n",
"plt.ylabel(\"Log Weekly Earnings\");"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"q4 parameter estimate:, 0.008603484260139599\n",
"q4 p-value:, 0.001494912718366745\n"
]
}
],
"source": [
"reduced_form = smf.ols(\n",
" \"log_wage ~ C(year_of_birth) + C(state_of_birth) + q4\",\n",
" data=factor_data).fit()\n",
"\n",
"print(\"q4 parameter estimate:, \", reduced_form.params[\"q4\"])\n",
"print(\"q4 p-value:, \", reduced_form.pvalues[\"q4\"])"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Estimating the average causal effect\n",
"\n",
"#### Option 1. Dividing reduced form by 1st stage.\n",
"\n",
"We can use the two regressions above - namely, the first stage:\n",
"```\n",
"first_stage = smf.ols(\n",
" 'years_of_schooling ~ C(year_of_birth) + C(state_of_birth) + q4',\n",
" data=factor_data).fit()\n",
"```\n",
"\n",
"And the reduced form:\n",
"```\n",
"reduced_form = smf.ols(\n",
" \"log_wage ~ C(year_of_birth) + C(state_of_birth) + q4\",\n",
" data=factor_data).fit()\n",
"```\n",
"\n",
"To get an unbiased IV estimate of the average causal effect, we can scale the effect of the reduced form coefficient by the first stage coefficient:\n",
"\n",
"$\n",
"\\mathrm{ATE}_{IV} = \\dfrac{\\text{Reduced Form}}{\\text{1st Stage}} \n",
"$\n",
"\n",
"Below, we see 0.08... which means we expect each additional year in school to increase wages by 8%."
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"0.08530286492084817"
]
},
"execution_count": 10,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"reduced_form.params[\"q4\"] / first_stage.params[\"q4\"]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Option 2. 2 stages least squares (2SLS)\n",
"\n",
"In 2SLS, we do the first stage as before -\n",
"\n",
"```\n",
"first_stage = smf.ols(\n",
" 'years_of_schooling ~ C(year_of_birth) + C(state_of_birth) + q4',\n",
" data=factor_data).fit()\n",
"```\n",
"\n",
"But then, run second stage where replace treatment variable with fitted values of the first stage. These are essentially \"the treatment purged from omitted variable bias\". Fitted values are the predicted values of the outcome variable from the model, when you input the values of the predictors into the model.\n",
"\n",
"We see the same results as with option 1."
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"0.08530286492089094"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"sls_data = factor_data.assign(\n",
" years_of_schooling_fitted=first_stage.fittedvalues)\n",
"\n",
"iv_by_hand = smf.ols(\n",
" \"log_wage ~ C(year_of_birth) + C(state_of_birth) + years_of_schooling_fitted\",\n",
" data=sls_data).fit()\n",
"\n",
"iv_by_hand.params[\"years_of_schooling_fitted\"]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Option 3. Automated 2SLS\n",
"\n",
"This is recommended, as else standard errors from second stage are a bit off."
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Parameter: 0.08530286494062492\n",
"SE: 0.025540812815019267\n",
"95 CI: [0.03524287 0.13536286]\n",
"P-value: 0.0008381914642161536\n"
]
}
],
"source": [
"formula = 'log_wage ~ 1 + C(year_of_birth) + C(state_of_birth) + [years_of_schooling ~ q4]'\n",
"iv2sls = IV2SLS.from_formula(formula, factor_data).fit()\n",
"parse(iv2sls)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can also run with multiple instruments - here, all the quarters of birth..."
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Parameter: 0.1076937048813452\n",
"SE: 0.01955714901081754\n",
"95 CI: [0.06936169 0.14602572]\n",
"P-value: 3.657974700921329e-08\n"
]
}
],
"source": [
"formula = 'log_wage ~ 1 + C(year_of_birth) + C(state_of_birth) + [years_of_schooling ~ q1+q2+q3]'\n",
"iv_many_zs = IV2SLS.from_formula(formula, factor_data).fit()\n",
"parse(iv_many_zs)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Summary\n",
"\n",
"Comparing traditional OLS with 2SLS:\n",
"* Education estimate lower with OLS than 2SLS, suggesting omitted variable bias may be less strong than thought\n",
"* 2SLS has much wider confidence intervals than OLS\n",
"\n",
"Some things to consider:\n",
"* If an instrumental variable has a **small correlation** with the treatment, then it is a weak instrument, and we won't be able to learn much from the instrument.\n",
"* 2SLS is biased towards OLS (i.e. if OLS has positive/negative bias, then so will 2SLS)\n",
"* Bias will increase with the number of instruments added\n",
"\n",
"Common mistakes:\n",
"* Doing IV by hand\n",
"* Using alternatives to OLS for the first stage. 'The consistency of IV relies on a property that only OLS can give, which is the orthogonality of the residuals, so anything different than OLS on the 1st stage will yield something biased.' "
]
}
],
"metadata": {
"kernelspec": {
"display_name": "venv_birth_outcomes_book",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.10.12"
}
},
"nbformat": 4,
"nbformat_minor": 2
}