examples/germany.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Comparative Politics and the Synthetic Control Method\n",
    "### Abadie, Diamond and Hainmueller"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This notebook reproduces the weights from the 2015 paper \"Comparative Politics and the Synthetic Control Method\" by Alberto Abadie, Alexis Diamond and Jens Hainmueller. The study data is contained in the file `../data/germany.csv` (more detailed information on this dataset is available in the appendix of the cited paper)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from pysyncon import Dataprep, Synth"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The study is carried out in two steps. In each step we prepare the study data using a `Dataprep` object that describes all the details needed to define the synthetic control study. This works similarly to the `dataprep` method in the `R` package `synth`.\n",
    "\n",
    "In the first run, the optimisation is carried out over the period 1981 to 1991, and the $V$ matrix obtained in this optimisation is then used in the second run, where the optimisation is carried out over the period 1960 to 1990. This serves to enforce that the predictor importances used in the final optimisation are those of the eighties.\n",
    "\n",
    "In each case, we supply the `Dataprep` object to a `Synth` object. In the second run, we can provide a custom $V$ matrix with the `custom_V` option.\n",
    "\n",
    "(For an explanation of each of the `Dataprep` arguments, see the package [documentation](https://sdfordham.github.io/pysyncon/dataprep.html#pysyncon.Dataprep))."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataprep\n",
      "Treated unit: West Germany\n",
      "Dependent variable: gdp\n",
      "Control units: USA, UK, Austria, Belgium, Denmark, France, Italy, Netherlands, Norway, Switzerland, Japan, Greece, Portugal, Spain, Australia, New Zealand\n",
      "Time range in data: 1960 - 2003\n",
      "Time range for loss minimization: range(1981, 1991)\n",
      "Time range for predictors: range(1971, 1981)\n",
      "Predictors: gdp, trade, infrate\n",
      "Special predictors:\n",
      "    `industry` over `range(1971, 1981)` using `mean`\n",
      "    `schooling` over `[1970, 1975]` using `mean`\n",
      "    `invest70` over `[1980]` using `mean`\n",
      "\n"
     ]
    }
   ],
   "source": [
    "df = pd.read_csv(\"../data/germany.csv\")\n",
    "\n",
    "dataprep_train = Dataprep(\n",
    "    foo=df,\n",
    "    predictors=[\"gdp\", \"trade\", \"infrate\"],\n",
    "    predictors_op=\"mean\",\n",
    "    time_predictors_prior=range(1971, 1981),\n",
    "    special_predictors=[\n",
    "        (\"industry\", range(1971, 1981), \"mean\"),\n",
    "        (\"schooling\", [1970, 1975], \"mean\"),\n",
    "        (\"invest70\", [1980], \"mean\"),\n",
    "    ],\n",
    "    dependent=\"gdp\",\n",
    "    unit_variable=\"country\",\n",
    "    time_variable=\"year\",\n",
    "    treatment_identifier=\"West Germany\",\n",
    "    controls_identifier=[\n",
    "        \"USA\",\n",
    "        \"UK\",\n",
    "        \"Austria\",\n",
    "        \"Belgium\",\n",
    "        \"Denmark\",\n",
    "        \"France\",\n",
    "        \"Italy\",\n",
    "        \"Netherlands\",\n",
    "        \"Norway\",\n",
    "        \"Switzerland\",\n",
    "        \"Japan\",\n",
    "        \"Greece\",\n",
    "        \"Portugal\",\n",
    "        \"Spain\",\n",
    "        \"Australia\",\n",
    "        \"New Zealand\",\n",
    "    ],\n",
    "    time_optimize_ssr=range(1981, 1991),\n",
    ")\n",
    "\n",
    "print(dataprep_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "synth_train = Synth()\n",
    "synth_train.fit(dataprep=dataprep_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataprep\n",
      "Treated unit: West Germany\n",
      "Dependent variable: gdp\n",
      "Control units: USA, UK, Austria, Belgium, Denmark, France, Italy, Netherlands, Norway, Switzerland, Japan, Greece, Portugal, Spain, Australia, New Zealand\n",
      "Time range in data: 1960 - 2003\n",
      "Time range for loss minimization: range(1960, 1990)\n",
      "Time range for predictors: range(1981, 1991)\n",
      "Predictors: gdp, trade, infrate\n",
      "Special predictors:\n",
      "    `industry` over `range(1981, 1991)` using `mean`\n",
      "    `schooling` over `[1980, 1985]` using `mean`\n",
      "    `invest80` over `[1980]` using `mean`\n",
      "\n"
     ]
    }
   ],
   "source": [
    "dataprep = Dataprep(\n",
    "    foo=df,\n",
    "    predictors=[\"gdp\", \"trade\", \"infrate\"],\n",
    "    predictors_op=\"mean\",\n",
    "    time_predictors_prior=range(1981, 1991),\n",
    "    special_predictors=[\n",
    "        (\"industry\", range(1981, 1991), \"mean\"),\n",
    "        (\"schooling\", [1980, 1985], \"mean\"),\n",
    "        (\"invest80\", [1980], \"mean\"),\n",
    "    ],\n",
    "    dependent=\"gdp\",\n",
    "    unit_variable=\"country\",\n",
    "    time_variable=\"year\",\n",
    "    treatment_identifier=\"West Germany\",\n",
    "    controls_identifier=[\n",
    "        \"USA\",\n",
    "        \"UK\",\n",
    "        \"Austria\",\n",
    "        \"Belgium\",\n",
    "        \"Denmark\",\n",
    "        \"France\",\n",
    "        \"Italy\",\n",
    "        \"Netherlands\",\n",
    "        \"Norway\",\n",
    "        \"Switzerland\",\n",
    "        \"Japan\",\n",
    "        \"Greece\",\n",
    "        \"Portugal\",\n",
    "        \"Spain\",\n",
    "        \"Australia\",\n",
    "        \"New Zealand\",\n",
    "    ],\n",
    "    time_optimize_ssr=range(1960, 1990),\n",
    ")\n",
    "\n",
    "print(dataprep)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "USA            0.216\n",
       "UK             0.000\n",
       "Austria        0.415\n",
       "Belgium        0.000\n",
       "Denmark        0.000\n",
       "France         0.000\n",
       "Italy          0.000\n",
       "Netherlands    0.098\n",
       "Norway         0.000\n",
       "Switzerland    0.108\n",
       "Japan          0.163\n",
       "Greece         0.000\n",
       "Portugal       0.000\n",
       "Spain          0.000\n",
       "Australia      0.000\n",
       "New Zealand    0.000\n",
       "Name: weights, dtype: float64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "synth = Synth()\n",
    "synth.fit(dataprep=dataprep, custom_V=synth_train.V)\n",
    "\n",
    "synth.weights()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The synthetic control obtained from the optimisation is: $$\\text{Synthetic Control} = 0.216 \\times \\text{USA} + 0.415 \\times \\text{Austria} + 0.098 \\times \\text{Nederlands} + 0.108 \\times \\text{Switzerland} + 0.163 \\times \\text{Japan}.$$"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The `path_plot` method shows the path of the treated unit and the synthetic control over time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "synth.path_plot(time_period=range(1960, 2004), treatment_time=1990)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The `gaps_plot` method shows the gaps (the difference between the treated unit and the synthetic control) over time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "synth.gaps_plot(time_period=range(1960, 2004), treatment_time=1990)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The summary function give more information on the predictor values. The first column shows the value of the $V$ matrix for each predictor, the column 'treated' shows the mean value of each predictor for the treated unit over the time period `time_predictors_prior`, the column 'synthetic' shows the mean value of each predictor for the synthetic control over the time period `time_predictors_prior` and finally the column 'sample mean' shows the sample mean of that predictor for all control units over the time period `time_predictors_prior` i.e. this is the same as the synthetic control with all weights equal."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>V</th>\n",
       "      <th>treated</th>\n",
       "      <th>synthetic</th>\n",
       "      <th>sample mean</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>gdp</th>\n",
       "      <td>0.240</td>\n",
       "      <td>15808.900</td>\n",
       "      <td>15777.800</td>\n",
       "      <td>13669.381</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>trade</th>\n",
       "      <td>0.205</td>\n",
       "      <td>56.778</td>\n",
       "      <td>56.980</td>\n",
       "      <td>59.831</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>infrate</th>\n",
       "      <td>0.188</td>\n",
       "      <td>2.595</td>\n",
       "      <td>3.441</td>\n",
       "      <td>7.617</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>special.1.industry</th>\n",
       "      <td>0.001</td>\n",
       "      <td>34.538</td>\n",
       "      <td>34.436</td>\n",
       "      <td>33.794</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>special.2.schooling</th>\n",
       "      <td>0.208</td>\n",
       "      <td>55.500</td>\n",
       "      <td>55.103</td>\n",
       "      <td>38.659</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>special.3.invest80</th>\n",
       "      <td>0.158</td>\n",
       "      <td>27.018</td>\n",
       "      <td>27.074</td>\n",
       "      <td>25.895</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         V    treated  synthetic  sample mean\n",
       "gdp                  0.240  15808.900  15777.800    13669.381\n",
       "trade                0.205     56.778     56.980       59.831\n",
       "infrate              0.188      2.595      3.441        7.617\n",
       "special.1.industry   0.001     34.538     34.436       33.794\n",
       "special.2.schooling  0.208     55.500     55.103       38.659\n",
       "special.3.invest80   0.158     27.018     27.074       25.895"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "synth.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Compute the average treatment effect on the treated unit (ATT) over the post-treatment time period. This method returns a standard error also."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'att': -1555.1346777620479, 'se': 317.6469306023242}"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "synth.att(time_period=range(1990, 2004))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Calculate 95% confidence intervals for the treatment effect for time periods $t=1991, \\ldots, 2000$. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>value</th>\n",
       "      <th>lower_ci</th>\n",
       "      <th>upper_ci</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>time</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1991</th>\n",
       "      <td>279.096860</td>\n",
       "      <td>43.070748</td>\n",
       "      <td>515.122971</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1992</th>\n",
       "      <td>99.762034</td>\n",
       "      <td>-136.264077</td>\n",
       "      <td>335.788146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1993</th>\n",
       "      <td>-631.543723</td>\n",
       "      <td>-867.569835</td>\n",
       "      <td>-395.517612</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1994</th>\n",
       "      <td>-1050.267990</td>\n",
       "      <td>-1286.294102</td>\n",
       "      <td>-814.241878</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1995</th>\n",
       "      <td>-1205.254923</td>\n",
       "      <td>-1441.281034</td>\n",
       "      <td>-969.228811</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1996</th>\n",
       "      <td>-1467.249163</td>\n",
       "      <td>-1703.275274</td>\n",
       "      <td>-1231.223051</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1997</th>\n",
       "      <td>-1954.374169</td>\n",
       "      <td>-2190.400281</td>\n",
       "      <td>-1718.348057</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1998</th>\n",
       "      <td>-2008.396030</td>\n",
       "      <td>-2244.422142</td>\n",
       "      <td>-1772.369918</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1999</th>\n",
       "      <td>-2160.627037</td>\n",
       "      <td>-2396.653148</td>\n",
       "      <td>-1924.600925</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000</th>\n",
       "      <td>-2620.733091</td>\n",
       "      <td>-2856.759203</td>\n",
       "      <td>-2384.706979</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            value     lower_ci     upper_ci\n",
       "time                                       \n",
       "1991   279.096860    43.070748   515.122971\n",
       "1992    99.762034  -136.264077   335.788146\n",
       "1993  -631.543723  -867.569835  -395.517612\n",
       "1994 -1050.267990 -1286.294102  -814.241878\n",
       "1995 -1205.254923 -1441.281034  -969.228811\n",
       "1996 -1467.249163 -1703.275274 -1231.223051\n",
       "1997 -1954.374169 -2190.400281 -1718.348057\n",
       "1998 -2008.396030 -2244.422142 -1772.369918\n",
       "1999 -2160.627037 -2396.653148 -1924.600925\n",
       "2000 -2620.733091 -2856.759203 -2384.706979"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "synth.confidence_interval(\n",
    "    alpha=0.05,\n",
    "    time_periods=[1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000],\n",
    "    custom_V=synth_train.V,\n",
    "    tol=0.01,\n",
    "    verbose=False,\n",
    ")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv",
   "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": "3.9.0"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "0affce9ccf703bfc26f5559d4b07023a564b817a32e04b2ea45527debe8fe1b9"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}