notebooks/analyze-filing-counts.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "05ac5b4b",
   "metadata": {},
   "source": [
    "# Estimating the national rate of guardianship filings\n",
    "\n",
    "This notebook attempts to estimate the number of guardianship cases filed annually, using data BuzzFeed News requested from all 50 US states and the District of Columbia. Data was not available from all states, and other states' data proved to be unusable, so we explore various approaches to extrapolating the national number from a limited sample. Please see this repository's landing page and/or `README.md` document for important details and methodological context."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "11bfa1f7",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:19.418786Z",
     "iopub.status.busy": "2021-09-15T18:32:19.410586Z",
     "iopub.status.idle": "2021-09-15T18:32:26.424549Z",
     "shell.execute_reply": "2021-09-15T18:32:26.425258Z"
    }
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import random\n",
    "from us import states\n",
    "import statsmodels.api as sm\n",
    "from scipy.stats import nbinom\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "sns.set()\n",
    "seed=sum(map(ord, \"BuzzFeed News\"))\n",
    "random.seed(seed)\n",
    "np.random.seed(seed)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "42693e91",
   "metadata": {},
   "source": [
    "# Load the guardianship data\n",
    "\n",
    "Here we load the guardianship filing counts we collected, for all years and disaggregated into several categories. See the `README.md` document for variable definitions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "c515d7f1",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:26.431141Z",
     "iopub.status.busy": "2021-09-15T18:32:26.429857Z",
     "iopub.status.idle": "2021-09-15T18:32:26.446552Z",
     "shell.execute_reply": "2021-09-15T18:32:26.446044Z"
    }
   },
   "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>state</th>\n",
       "      <th>year</th>\n",
       "      <th>case_action</th>\n",
       "      <th>case_type</th>\n",
       "      <th>age_group</th>\n",
       "      <th>count</th>\n",
       "      <th>for_total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Alaska</td>\n",
       "      <td>2020</td>\n",
       "      <td>Ambiguous</td>\n",
       "      <td>G+C+Joint</td>\n",
       "      <td>Adults+Minors</td>\n",
       "      <td>1171</td>\n",
       "      <td>Y</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Alaska</td>\n",
       "      <td>2019</td>\n",
       "      <td>Ambiguous</td>\n",
       "      <td>G+C+Joint</td>\n",
       "      <td>Adults+Minors</td>\n",
       "      <td>1084</td>\n",
       "      <td>Y</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Alaska</td>\n",
       "      <td>2018</td>\n",
       "      <td>Ambiguous</td>\n",
       "      <td>G+C+Joint</td>\n",
       "      <td>Adults+Minors</td>\n",
       "      <td>1227</td>\n",
       "      <td>Y</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Alaska</td>\n",
       "      <td>2017</td>\n",
       "      <td>Ambiguous</td>\n",
       "      <td>G+C+Joint</td>\n",
       "      <td>Adults+Minors</td>\n",
       "      <td>1090</td>\n",
       "      <td>Y</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Arizona</td>\n",
       "      <td>2020</td>\n",
       "      <td>Ambiguous</td>\n",
       "      <td>Guardianship ONLY</td>\n",
       "      <td>Adults ONLY</td>\n",
       "      <td>2520</td>\n",
       "      <td>Y</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     state  year case_action          case_type      age_group  count  \\\n",
       "0   Alaska  2020   Ambiguous          G+C+Joint  Adults+Minors   1171   \n",
       "1   Alaska  2019   Ambiguous          G+C+Joint  Adults+Minors   1084   \n",
       "2   Alaska  2018   Ambiguous          G+C+Joint  Adults+Minors   1227   \n",
       "3   Alaska  2017   Ambiguous          G+C+Joint  Adults+Minors   1090   \n",
       "4  Arizona  2020   Ambiguous  Guardianship ONLY    Adults ONLY   2520   \n",
       "\n",
       "  for_total  \n",
       "0         Y  \n",
       "1         Y  \n",
       "2         Y  \n",
       "3         Y  \n",
       "4         Y  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counts = pd.read_csv(\"../data/raw/filing-counts.csv\")\n",
    "counts.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "6c3fbcbc",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:26.451359Z",
     "iopub.status.busy": "2021-09-15T18:32:26.450695Z",
     "iopub.status.idle": "2021-09-15T18:32:26.453092Z",
     "shell.execute_reply": "2021-09-15T18:32:26.453514Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "380"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(counts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "40f9df14",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:26.461370Z",
     "iopub.status.busy": "2021-09-15T18:32:26.460554Z",
     "iopub.status.idle": "2021-09-15T18:32:26.463778Z",
     "shell.execute_reply": "2021-09-15T18:32:26.464335Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "37"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counts[\"state\"].nunique()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "15a41ed8",
   "metadata": {},
   "source": [
    "Although we've collected data from most states, only 23 jurisdictions (22 states plus the District of Columbia) have data fully usable for generating annual totals of *adult* guardianship filings. See the `README.md` document for more details."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "eaa4c493",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:26.473308Z",
     "iopub.status.busy": "2021-09-15T18:32:26.472505Z",
     "iopub.status.idle": "2021-09-15T18:32:26.475313Z",
     "shell.execute_reply": "2021-09-15T18:32:26.475800Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "23"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    counts\n",
    "    .loc[lambda df: df[\"for_total\"] == \"Y\"]\n",
    "    .loc[lambda df: df[\"age_group\"] == \"Adults ONLY\"]\n",
    "    [\"state\"]\n",
    "    .nunique()\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4ba4fa63",
   "metadata": {},
   "source": [
    "# Examine years available\n",
    "\n",
    "All usable jurisdictions have data for 2019, and all but one for 2018. (And although we have collected data for 2020 for many states, we will generally ignore it due to the likelihood that the COVID-19 pandemic affected filing rates.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "90ad44ef",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:26.485090Z",
     "iopub.status.busy": "2021-09-15T18:32:26.484022Z",
     "iopub.status.idle": "2021-09-15T18:32:26.490179Z",
     "shell.execute_reply": "2021-09-15T18:32:26.490618Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "year\n",
       "2016     7\n",
       "2017    14\n",
       "2018    22\n",
       "2019    23\n",
       "2020    19\n",
       "Name: state, dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    counts\n",
    "    .loc[lambda df: df[\"for_total\"] == \"Y\"]\n",
    "    .loc[lambda df: df[\"age_group\"] == \"Adults ONLY\"]\n",
    "    .groupby([\"year\"])\n",
    "    [\"state\"]\n",
    "    .nunique()\n",
    "    .sort_index()\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "74b7c3fc",
   "metadata": {},
   "source": [
    "Here we identify the latest year of usable data available for each jurisdiction. In this case, all values are `2019`, but the approach implemented in this code allows for flexibility when/if the data changes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "d52a13b4",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:26.505656Z",
     "iopub.status.busy": "2021-09-15T18:32:26.501041Z",
     "iopub.status.idle": "2021-09-15T18:32:26.508609Z",
     "shell.execute_reply": "2021-09-15T18:32:26.509035Z"
    }
   },
   "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>state</th>\n",
       "      <th>year_max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Arizona</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Arkansas</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Colorado</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Connecticut</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>District of Columbia</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Idaho</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Kansas</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Kentucky</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Massachusetts</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Michigan</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Minnesota</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Missouri</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Nebraska</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>New Hampshire</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>New Mexico</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>North Dakota</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>Ohio</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>Oregon</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>Pennsylvania</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>Utah</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>Vermont</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>Virginia</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>Washington</td>\n",
       "      <td>2019</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   state  year_max\n",
       "0                Arizona      2019\n",
       "1               Arkansas      2019\n",
       "2               Colorado      2019\n",
       "3            Connecticut      2019\n",
       "4   District of Columbia      2019\n",
       "5                  Idaho      2019\n",
       "6                 Kansas      2019\n",
       "7               Kentucky      2019\n",
       "8          Massachusetts      2019\n",
       "9               Michigan      2019\n",
       "10             Minnesota      2019\n",
       "11              Missouri      2019\n",
       "12              Nebraska      2019\n",
       "13         New Hampshire      2019\n",
       "14            New Mexico      2019\n",
       "15          North Dakota      2019\n",
       "16                  Ohio      2019\n",
       "17                Oregon      2019\n",
       "18          Pennsylvania      2019\n",
       "19                  Utah      2019\n",
       "20               Vermont      2019\n",
       "21              Virginia      2019\n",
       "22            Washington      2019"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "states_latest_year = (\n",
    "    counts\n",
    "    .loc[lambda df: df[\"for_total\"] == \"Y\"]\n",
    "    .loc[lambda df: df[\"age_group\"] == \"Adults ONLY\"]\n",
    "    .loc[lambda df: df[\"year\"] < 2020]\n",
    "    .groupby([\"state\"])\n",
    "    [\"year\"].max()\n",
    "    .to_frame(\"year_max\")\n",
    "    .reset_index()\n",
    ")\n",
    "\n",
    "states_latest_year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "8d99c09c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:26.517062Z",
     "iopub.status.busy": "2021-09-15T18:32:26.516188Z",
     "iopub.status.idle": "2021-09-15T18:32:26.519075Z",
     "shell.execute_reply": "2021-09-15T18:32:26.519494Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2019    23\n",
       "Name: year_max, dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "states_latest_year[\"year_max\"].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4409a3eb",
   "metadata": {},
   "source": [
    "Using the table above, we select *just* the counts belonging to each jurisdiction's most recent year:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "393aa91e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:26.542251Z",
     "iopub.status.busy": "2021-09-15T18:32:26.541516Z",
     "iopub.status.idle": "2021-09-15T18:32:26.544331Z",
     "shell.execute_reply": "2021-09-15T18:32:26.544751Z"
    }
   },
   "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>state</th>\n",
       "      <th>year</th>\n",
       "      <th>case_action</th>\n",
       "      <th>case_type</th>\n",
       "      <th>age_group</th>\n",
       "      <th>count</th>\n",
       "      <th>for_total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Arizona</td>\n",
       "      <td>2019</td>\n",
       "      <td>Ambiguous</td>\n",
       "      <td>Guardianship ONLY</td>\n",
       "      <td>Adults ONLY</td>\n",
       "      <td>2383</td>\n",
       "      <td>Y</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Arizona</td>\n",
       "      <td>2019</td>\n",
       "      <td>Ambiguous</td>\n",
       "      <td>Guardianship ONLY</td>\n",
       "      <td>Minors ONLY</td>\n",
       "      <td>3608</td>\n",
       "      <td>N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Arizona</td>\n",
       "      <td>2019</td>\n",
       "      <td>Ambiguous</td>\n",
       "      <td>Conservatorship ONLY</td>\n",
       "      <td>Adults ONLY</td>\n",
       "      <td>4003</td>\n",
       "      <td>N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Arizona</td>\n",
       "      <td>2019</td>\n",
       "      <td>Ambiguous</td>\n",
       "      <td>Conservatorship ONLY</td>\n",
       "      <td>Minors ONLY</td>\n",
       "      <td>1334</td>\n",
       "      <td>N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>Arkansas</td>\n",
       "      <td>2019</td>\n",
       "      <td>Ambiguous</td>\n",
       "      <td>G+C+Joint</td>\n",
       "      <td>Adults ONLY</td>\n",
       "      <td>1164</td>\n",
       "      <td>Y</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       state  year case_action             case_type    age_group  count  \\\n",
       "8    Arizona  2019   Ambiguous     Guardianship ONLY  Adults ONLY   2383   \n",
       "9    Arizona  2019   Ambiguous     Guardianship ONLY  Minors ONLY   3608   \n",
       "10   Arizona  2019   Ambiguous  Conservatorship ONLY  Adults ONLY   4003   \n",
       "11   Arizona  2019   Ambiguous  Conservatorship ONLY  Minors ONLY   1334   \n",
       "19  Arkansas  2019   Ambiguous             G+C+Joint  Adults ONLY   1164   \n",
       "\n",
       "   for_total  \n",
       "8          Y  \n",
       "9          N  \n",
       "10         N  \n",
       "11         N  \n",
       "19         Y  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "latest = (\n",
    "    counts\n",
    "    .pipe(lambda df: (\n",
    "        df\n",
    "        .merge(\n",
    "            states_latest_year,\n",
    "            how = \"left\",\n",
    "            on = \"state\",\n",
    "        )\n",
    "    ))\n",
    "    .loc[lambda df: df[\"year\"] == df[\"year_max\"]]\n",
    "    .drop(columns = [\"year_max\"])\n",
    ")\n",
    "\n",
    "latest.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "e59bd4c5",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:26.549221Z",
     "iopub.status.busy": "2021-09-15T18:32:26.548460Z",
     "iopub.status.idle": "2021-09-15T18:32:26.553625Z",
     "shell.execute_reply": "2021-09-15T18:32:26.554110Z"
    }
   },
   "outputs": [],
   "source": [
    "latest.to_csv(\"../data/processed/latest-counts.csv\", index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3df68181",
   "metadata": {},
   "source": [
    "# Load population data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "815f079c",
   "metadata": {},
   "source": [
    "In order to get per-capita rates, we'll need popuation data. Here we use the [Census Population Estimates Program's estimates for 2019](https://www2.census.gov/programs-surveys/popest/tables/2010-2019/state/detail/SCPRC-EST2019-18+POP-RES.xlsx), overall and for age 18+:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "36523615",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:26.559926Z",
     "iopub.status.busy": "2021-09-15T18:32:26.559193Z",
     "iopub.status.idle": "2021-09-15T18:32:27.209156Z",
     "shell.execute_reply": "2021-09-15T18:32:27.209790Z"
    }
   },
   "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>state</th>\n",
       "      <th>pop_total</th>\n",
       "      <th>pop_adult</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Alabama</td>\n",
       "      <td>4903185</td>\n",
       "      <td>3814879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Alaska</td>\n",
       "      <td>731545</td>\n",
       "      <td>551562</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Arizona</td>\n",
       "      <td>7278717</td>\n",
       "      <td>5638481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Arkansas</td>\n",
       "      <td>3017804</td>\n",
       "      <td>2317649</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>California</td>\n",
       "      <td>39512223</td>\n",
       "      <td>30617582</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Colorado</td>\n",
       "      <td>5758736</td>\n",
       "      <td>4499217</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Connecticut</td>\n",
       "      <td>3565287</td>\n",
       "      <td>2837847</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Delaware</td>\n",
       "      <td>973764</td>\n",
       "      <td>770192</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>District of Columbia</td>\n",
       "      <td>705749</td>\n",
       "      <td>577581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Florida</td>\n",
       "      <td>21477737</td>\n",
       "      <td>17247808</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  state  pop_total  pop_adult\n",
       "0               Alabama    4903185    3814879\n",
       "1                Alaska     731545     551562\n",
       "2               Arizona    7278717    5638481\n",
       "3              Arkansas    3017804    2317649\n",
       "4            California   39512223   30617582\n",
       "5              Colorado    5758736    4499217\n",
       "6           Connecticut    3565287    2837847\n",
       "7              Delaware     973764     770192\n",
       "8  District of Columbia     705749     577581\n",
       "9               Florida   21477737   17247808"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "populations = (\n",
    "    pd.read_excel(\n",
    "        \"../data/raw/census-populations-2019.xlsx\",\n",
    "        header=None,\n",
    "        skiprows=9,\n",
    "        skipfooter=7,\n",
    "        names = [ \"state\", \"pop_total\", \"pop_adult\", \"pct\" ],\n",
    "        dtype = {\n",
    "            \"pop_total\": int,\n",
    "            \"pop_adult\": int,\n",
    "        }\n",
    "    )\n",
    "    .drop(columns = [ \"pct\" ])\n",
    "    .assign(\n",
    "        state = lambda df: df[\"state\"].str.replace(r\"^\\.\", \"\", regex=True),\n",
    "    )\n",
    ")\n",
    "\n",
    "populations.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "77bfd124",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:27.215225Z",
     "iopub.status.busy": "2021-09-15T18:32:27.214421Z",
     "iopub.status.idle": "2021-09-15T18:32:27.217619Z",
     "shell.execute_reply": "2021-09-15T18:32:27.217121Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "51"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(populations)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "ba4c8e0d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:27.222226Z",
     "iopub.status.busy": "2021-09-15T18:32:27.221519Z",
     "iopub.status.idle": "2021-09-15T18:32:27.223224Z",
     "shell.execute_reply": "2021-09-15T18:32:27.223837Z"
    }
   },
   "outputs": [],
   "source": [
    "# Check that our state names match the Census's\n",
    "assert len(set(counts[\"state\"]) - set(populations[\"state\"])) == 0"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "77e074b6",
   "metadata": {},
   "source": [
    "# Total the total-able data, and calculate per capita rates\n",
    "\n",
    "We merge the adult populations with the case count data and, for each jurisdiction and year, sum all `count`s where `for_total` is `\"Y\"` and `age_group` is `Adults ONLY`. Then we calculate the rate of filings per 100,000 adult residents."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "b1c80475",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:27.229567Z",
     "iopub.status.busy": "2021-09-15T18:32:27.228917Z",
     "iopub.status.idle": "2021-09-15T18:32:27.231342Z",
     "shell.execute_reply": "2021-09-15T18:32:27.231797Z"
    }
   },
   "outputs": [],
   "source": [
    "def get_totals_and_rates(subset):\n",
    "    return (\n",
    "        subset\n",
    "        .groupby([\"state\", \"year\"])\n",
    "        [\"count\"]\n",
    "        .sum()\n",
    "        .to_frame(\"total\")\n",
    "        .reset_index()\n",
    "        .merge(\n",
    "            populations\n",
    "            [[\n",
    "                \"state\",\n",
    "                \"pop_adult\",\n",
    "            ]],\n",
    "            how = \"left\",\n",
    "            on = \"state\"\n",
    "        )\n",
    "        .assign(\n",
    "            per_100k = lambda df: df[\"total\"] / (df[\"pop_adult\"] / 1e5)\n",
    "        )\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "144b5f4c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:27.248053Z",
     "iopub.status.busy": "2021-09-15T18:32:27.246135Z",
     "iopub.status.idle": "2021-09-15T18:32:27.268436Z",
     "shell.execute_reply": "2021-09-15T18:32:27.269136Z"
    }
   },
   "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>state</th>\n",
       "      <th>year</th>\n",
       "      <th>total</th>\n",
       "      <th>pop_adult</th>\n",
       "      <th>per_100k</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Connecticut</td>\n",
       "      <td>2019</td>\n",
       "      <td>4700</td>\n",
       "      <td>2837847</td>\n",
       "      <td>165.618513</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Michigan</td>\n",
       "      <td>2019</td>\n",
       "      <td>12655</td>\n",
       "      <td>7842924</td>\n",
       "      <td>161.355637</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Nebraska</td>\n",
       "      <td>2019</td>\n",
       "      <td>2109</td>\n",
       "      <td>1458334</td>\n",
       "      <td>144.617077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>Vermont</td>\n",
       "      <td>2019</td>\n",
       "      <td>512</td>\n",
       "      <td>509984</td>\n",
       "      <td>100.395307</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>District of Columbia</td>\n",
       "      <td>2019</td>\n",
       "      <td>575</td>\n",
       "      <td>577581</td>\n",
       "      <td>99.553136</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>New Hampshire</td>\n",
       "      <td>2019</td>\n",
       "      <td>945</td>\n",
       "      <td>1104458</td>\n",
       "      <td>85.562330</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>Ohio</td>\n",
       "      <td>2019</td>\n",
       "      <td>6991</td>\n",
       "      <td>9111081</td>\n",
       "      <td>76.730741</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Massachusetts</td>\n",
       "      <td>2019</td>\n",
       "      <td>4059</td>\n",
       "      <td>5539703</td>\n",
       "      <td>73.271076</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Missouri</td>\n",
       "      <td>2019</td>\n",
       "      <td>2922</td>\n",
       "      <td>4766843</td>\n",
       "      <td>61.298432</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Minnesota</td>\n",
       "      <td>2019</td>\n",
       "      <td>2658</td>\n",
       "      <td>4336475</td>\n",
       "      <td>61.294023</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>North Dakota</td>\n",
       "      <td>2019</td>\n",
       "      <td>314</td>\n",
       "      <td>581891</td>\n",
       "      <td>53.961996</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Kansas</td>\n",
       "      <td>2019</td>\n",
       "      <td>1142</td>\n",
       "      <td>2213064</td>\n",
       "      <td>51.602665</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Arkansas</td>\n",
       "      <td>2019</td>\n",
       "      <td>1164</td>\n",
       "      <td>2317649</td>\n",
       "      <td>50.223308</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>Washington</td>\n",
       "      <td>2019</td>\n",
       "      <td>2987</td>\n",
       "      <td>5951832</td>\n",
       "      <td>50.186228</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Idaho</td>\n",
       "      <td>2019</td>\n",
       "      <td>594</td>\n",
       "      <td>1338864</td>\n",
       "      <td>44.365970</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Kentucky</td>\n",
       "      <td>2019</td>\n",
       "      <td>1493</td>\n",
       "      <td>3464802</td>\n",
       "      <td>43.090485</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>New Mexico</td>\n",
       "      <td>2019</td>\n",
       "      <td>696</td>\n",
       "      <td>1620991</td>\n",
       "      <td>42.936697</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Arizona</td>\n",
       "      <td>2019</td>\n",
       "      <td>2383</td>\n",
       "      <td>5638481</td>\n",
       "      <td>42.263156</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>Oregon</td>\n",
       "      <td>2019</td>\n",
       "      <td>1355</td>\n",
       "      <td>3351175</td>\n",
       "      <td>40.433579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>Virginia</td>\n",
       "      <td>2019</td>\n",
       "      <td>2518</td>\n",
       "      <td>6674671</td>\n",
       "      <td>37.724706</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>Utah</td>\n",
       "      <td>2019</td>\n",
       "      <td>823</td>\n",
       "      <td>2274774</td>\n",
       "      <td>36.179418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>Pennsylvania</td>\n",
       "      <td>2019</td>\n",
       "      <td>3304</td>\n",
       "      <td>10167376</td>\n",
       "      <td>32.496093</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Colorado</td>\n",
       "      <td>2019</td>\n",
       "      <td>1417</td>\n",
       "      <td>4499217</td>\n",
       "      <td>31.494369</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   state  year  total  pop_adult    per_100k\n",
       "3            Connecticut  2019   4700    2837847  165.618513\n",
       "9               Michigan  2019  12655    7842924  161.355637\n",
       "12              Nebraska  2019   2109    1458334  144.617077\n",
       "20               Vermont  2019    512     509984  100.395307\n",
       "4   District of Columbia  2019    575     577581   99.553136\n",
       "13         New Hampshire  2019    945    1104458   85.562330\n",
       "16                  Ohio  2019   6991    9111081   76.730741\n",
       "8          Massachusetts  2019   4059    5539703   73.271076\n",
       "11              Missouri  2019   2922    4766843   61.298432\n",
       "10             Minnesota  2019   2658    4336475   61.294023\n",
       "15          North Dakota  2019    314     581891   53.961996\n",
       "6                 Kansas  2019   1142    2213064   51.602665\n",
       "1               Arkansas  2019   1164    2317649   50.223308\n",
       "22            Washington  2019   2987    5951832   50.186228\n",
       "5                  Idaho  2019    594    1338864   44.365970\n",
       "7               Kentucky  2019   1493    3464802   43.090485\n",
       "14            New Mexico  2019    696    1620991   42.936697\n",
       "0                Arizona  2019   2383    5638481   42.263156\n",
       "17                Oregon  2019   1355    3351175   40.433579\n",
       "21              Virginia  2019   2518    6674671   37.724706\n",
       "19                  Utah  2019    823    2274774   36.179418\n",
       "18          Pennsylvania  2019   3304   10167376   32.496093\n",
       "2               Colorado  2019   1417    4499217   31.494369"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "latest_totals = get_totals_and_rates(\n",
    "    latest\n",
    "    .loc[lambda df: df[\"for_total\"] == \"Y\"]\n",
    "    .loc[lambda df: df[\"age_group\"] == \"Adults ONLY\"]\n",
    ")\n",
    "\n",
    "latest_totals.sort_values(\"per_100k\", ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "a4145225",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:27.273957Z",
     "iopub.status.busy": "2021-09-15T18:32:27.272859Z",
     "iopub.status.idle": "2021-09-15T18:32:27.276804Z",
     "shell.execute_reply": "2021-09-15T18:32:27.276244Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "23"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(latest_totals)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a903fcea",
   "metadata": {},
   "source": [
    "How many filings do these jurisdictions account for?:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "08b9ded4",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:27.282016Z",
     "iopub.status.busy": "2021-09-15T18:32:27.280830Z",
     "iopub.status.idle": "2021-09-15T18:32:27.283831Z",
     "shell.execute_reply": "2021-09-15T18:32:27.284257Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "58316\n"
     ]
    }
   ],
   "source": [
    "known_total = latest_totals[\"total\"].sum()\n",
    "print(known_total)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0154cac4",
   "metadata": {},
   "source": [
    "What proportion of the full adult population (of the 50 states and DC) do these jurisdictions account for?:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "5add95db",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:27.292931Z",
     "iopub.status.busy": "2021-09-15T18:32:27.292072Z",
     "iopub.status.idle": "2021-09-15T18:32:27.294598Z",
     "shell.execute_reply": "2021-09-15T18:32:27.295270Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.346"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    (\n",
    "        populations\n",
    "        .loc[lambda df: df[\"state\"].isin(latest_totals[\"state\"])]\n",
    "        [\"pop_adult\"]\n",
    "        .sum() / populations[\"pop_adult\"].sum()\n",
    "    ) \n",
    "    .round(3)   \n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6758a3cb",
   "metadata": {},
   "source": [
    "And here we isolate the states for which we don't have usable totals:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "e34e0726",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:27.301541Z",
     "iopub.status.busy": "2021-09-15T18:32:27.300793Z",
     "iopub.status.idle": "2021-09-15T18:32:27.303850Z",
     "shell.execute_reply": "2021-09-15T18:32:27.304367Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "28"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "not_counted = (\n",
    "    populations\n",
    "    .loc[~populations[\"state\"].isin(latest_totals[\"state\"])]\n",
    ")\n",
    "\n",
    "len(not_counted)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bf67c8a4",
   "metadata": {},
   "source": [
    "# Examine the distribution of per capita filing rates\n",
    "\n",
    "To get an overview of our data, we examine some basic metrics (mean, median, range, standard deviation) of the filing rates:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "c7d07457",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:27.311741Z",
     "iopub.status.busy": "2021-09-15T18:32:27.310920Z",
     "iopub.status.idle": "2021-09-15T18:32:27.316907Z",
     "shell.execute_reply": "2021-09-15T18:32:27.317330Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count     23.000000\n",
       "mean      68.984998\n",
       "std       40.165010\n",
       "min       31.494369\n",
       "25%       42.599926\n",
       "50%       51.602665\n",
       "75%       81.146536\n",
       "max      165.618513\n",
       "Name: per_100k, dtype: float64"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    latest_totals\n",
    "    [\"per_100k\"]\n",
    "    .describe()\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5dca5a45",
   "metadata": {},
   "source": [
    "... and create a histogram of the distribution:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "d1865f30",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:27.322159Z",
     "iopub.status.busy": "2021-09-15T18:32:27.321144Z",
     "iopub.status.idle": "2021-09-15T18:32:27.323258Z",
     "shell.execute_reply": "2021-09-15T18:32:27.323678Z"
    }
   },
   "outputs": [],
   "source": [
    "bin_width = 40\n",
    "bin_max = 250\n",
    "bin_count = int(bin_max / bin_width)\n",
    "bins = np.arange(0, bin_max, bin_width)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "b87a1e34",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:27.328315Z",
     "iopub.status.busy": "2021-09-15T18:32:27.327657Z",
     "iopub.status.idle": "2021-09-15T18:32:27.567725Z",
     "shell.execute_reply": "2021-09-15T18:32:27.568382Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = (\n",
    "    latest_totals\n",
    "    [\"per_100k\"]\n",
    "    .hist(bins=bins)\n",
    ")\n",
    "plt.xticks(bins)\n",
    "ax.set_xlabel(\"Cases per 100k Adults\")\n",
    "ax.set_ylabel(\"# States\")\n",
    "pass"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f4f65502",
   "metadata": {},
   "source": [
    "# Generating simple national estimates\n",
    "\n",
    "A simple approach to estimating the number of cases filed nationally in 2019 would be to multiply the *overall* filing rate per capita in the jurisdictions for which we have totals and to multiply it by the total adult population in all 50 states and DC (shortened below to the \"total US adult population,\" although please note that this number excludes US territories):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "9ddb85b7",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:27.573741Z",
     "iopub.status.busy": "2021-09-15T18:32:27.572957Z",
     "iopub.status.idle": "2021-09-15T18:32:27.576584Z",
     "shell.execute_reply": "2021-09-15T18:32:27.577031Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "168771"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "round(\n",
    "    (latest_totals[\"total\"].sum() / latest_totals[\"pop_adult\"].sum()) * \n",
    "    populations[\"pop_adult\"].sum() \n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "62a72018",
   "metadata": {},
   "source": [
    "Alternatively, but in the same vein, we could apply the *average state filing rate* to the total US adult population — an approach that mirrors [the methodology applied](https://www.eldersandcourts.org/__data/assets/pdf_file/0021/5844/ovc-brief-7.pdf) (see p. 4) by the National Center for State Courts to a survey of 2015 adult guardianship/conservatorship filings in 16 states, which resulted in a nearly identical figure (\"an estimated 180,000,\" [elsewhere summarized](http://web.archive.org/web/20170906012840/https://www.ncsc.org/Newsroom/Backgrounder/2016/Guardianship.aspx) as \"176,000\"):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "ed072a68",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:27.582179Z",
     "iopub.status.busy": "2021-09-15T18:32:27.581421Z",
     "iopub.status.idle": "2021-09-15T18:32:27.584170Z",
     "shell.execute_reply": "2021-09-15T18:32:27.584795Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "176050"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "round(\n",
    "    (latest_totals[\"per_100k\"].mean() * populations[\"pop_adult\"].sum() / 1e5))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "051604ef",
   "metadata": {},
   "source": [
    "... or apply that average rate *only to the remaining states*, which reduces the estimate slightly:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "4af79bd2",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:27.590570Z",
     "iopub.status.busy": "2021-09-15T18:32:27.589181Z",
     "iopub.status.idle": "2021-09-15T18:32:27.592426Z",
     "shell.execute_reply": "2021-09-15T18:32:27.592840Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "173535"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "round(\n",
    "    (latest_totals[\"per_100k\"].mean() * not_counted[\"pop_adult\"].sum() / 1e5)\n",
    "    + known_total\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "81314c66",
   "metadata": {},
   "source": [
    "These approaches may provide useful ballpark estimates but they also have some key disadvantages — mainly that they produce only single-number estimates, with no sense of how widely the range of actual possibilities may be.\n",
    "\n",
    "The first approach also gives more weight to the larger states we have observed, because they account for more of the population in the sample."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "812ceceb",
   "metadata": {},
   "source": [
    "# Bootstrapping a national estimate\n",
    "\n",
    "Another approach we can take, to incorporate the uncertainty involved in the extrapolations, is to \"bootstrap\" a distribution of estimates for the remaining states. This method hypothetically assigns each of those states to one of the per capita filing rates we have observed (selected at random), calculates what the total would be for each of those state-rate combinations, and then sums up the total. Finally, we add each hypothetical's total for the those states to our known total."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "7c5215c0",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:27.600181Z",
     "iopub.status.busy": "2021-09-15T18:32:27.599418Z",
     "iopub.status.idle": "2021-09-15T18:32:27.601299Z",
     "shell.execute_reply": "2021-09-15T18:32:27.601726Z"
    }
   },
   "outputs": [],
   "source": [
    "def gen_random_items(pool, n):\n",
    "    return [ random.choice(list(pool)) for _ in range(n) ]\n",
    "\n",
    "def gen_random_estimate(populations, rates):\n",
    "    rand_rates = gen_random_items(rates, len(populations))\n",
    "    return (populations * rand_rates).sum()\n",
    "\n",
    "def bootstrap(populations, rates, n=1e4):\n",
    "    return pd.Series(gen_random_estimate(populations, rates) for _ in range(int(n)))\n",
    "\n",
    "def describe(series):\n",
    "    return (\n",
    "        series\n",
    "        .describe(percentiles=[0.025, 0.25, 0.5, 0.75, 0.975])\n",
    "        .astype(int)\n",
    "    )\n",
    "\n",
    "def hist(series):\n",
    "    return (\n",
    "        series\n",
    "        .hist(\n",
    "            bins=np.arange(1e5, 3.1e5, int(1e4)),\n",
    "            alpha=0.5\n",
    "        )\n",
    "    )"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "925b0278",
   "metadata": {},
   "source": [
    "Basic statistics about the resulting range of estimates:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "28c8bb63",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:27.690959Z",
     "iopub.status.busy": "2021-09-15T18:32:27.650817Z",
     "iopub.status.idle": "2021-09-15T18:32:31.375464Z",
     "shell.execute_reply": "2021-09-15T18:32:31.376186Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count     10000\n",
       "mean     173618\n",
       "std       19294\n",
       "min      127230\n",
       "2.5%     142919\n",
       "25%      158982\n",
       "50%      171300\n",
       "75%      186030\n",
       "97.5%    216686\n",
       "max      249941\n",
       "dtype: int64"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bootstrapped_not_counted = bootstrap(\n",
    "    not_counted[\"pop_adult\"] / 1e5,\n",
    "    latest_totals[\"per_100k\"],\n",
    "    n = 1e4\n",
    ")\n",
    "\n",
    "describe(bootstrapped_not_counted + known_total)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "362ebe94",
   "metadata": {},
   "source": [
    "... and a histogram of of them:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "44ac5c00",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:31.398392Z",
     "iopub.status.busy": "2021-09-15T18:32:31.397780Z",
     "iopub.status.idle": "2021-09-15T18:32:31.641595Z",
     "shell.execute_reply": "2021-09-15T18:32:31.642324Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "hist(bootstrapped_not_counted + known_total);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "20182653",
   "metadata": {},
   "source": [
    "# Modeling the counts through the negative binomial distribution\n",
    "\n",
    "The data collected represents a finite number of observations. If we had infinitely many states' of data, what might this distribution look like? We can \"model\" this by fitting the observed data to a statistical distribution that aligns — conceptually and empirically — with what we've seen.\n",
    "\n",
    "A reasonable candidate is the [negative binomial distribution](https://en.wikipedia.org/wiki/Negative_binomial_distribution), which statisticians often use to model count-based metrics and which in this case appears to match our observations reasonably well. Below we estimate the parameters for our negative binomial distribution, based on our *observed* per capita rates' mean and variance, and overlay it on our observed distribution in two ways:\n",
    "\n",
    "- A smooth line representing the negative binomial's [probability mass function](https://en.wikipedia.org/wiki/Probability_mass_function)\n",
    "\n",
    "- A histogram representing 10,000 hypothetical observations randomly generated from the negative binomial"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "b147876d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:31.649878Z",
     "iopub.status.busy": "2021-09-15T18:32:31.648857Z",
     "iopub.status.idle": "2021-09-15T18:32:31.653732Z",
     "shell.execute_reply": "2021-09-15T18:32:31.654665Z"
    }
   },
   "outputs": [],
   "source": [
    "def calculate_nbinom_np(rates):\n",
    "    mean = rates.mean()\n",
    "    variance = rates.var()\n",
    "    n = pow(mean, 2) / (variance - mean)\n",
    "    p = mean / variance\n",
    "    return n, p\n",
    "\n",
    "def nbinom_from_rates(rates):\n",
    "    return nbinom(*calculate_nbinom_np(rates))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "8a94f12d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:31.661753Z",
     "iopub.status.busy": "2021-09-15T18:32:31.660822Z",
     "iopub.status.idle": "2021-09-15T18:32:31.664038Z",
     "shell.execute_reply": "2021-09-15T18:32:31.663288Z"
    }
   },
   "outputs": [],
   "source": [
    "nbinom_dist = nbinom_from_rates(latest_totals[\"per_100k\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "19c9c421",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:31.682407Z",
     "iopub.status.busy": "2021-09-15T18:32:31.680750Z",
     "iopub.status.idle": "2021-09-15T18:32:31.933696Z",
     "shell.execute_reply": "2021-09-15T18:32:31.934240Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "_x = np.arange(0, bin_max)\n",
    "\n",
    "ax = (\n",
    "    latest_totals\n",
    "    [\"per_100k\"]\n",
    "    .hist(bins=bins, density=True)\n",
    ")\n",
    "plt.xticks(bins)\n",
    "ax.set_xlabel(\"Cases per 100k Adults\")\n",
    "ax.set_ylabel(\"Observation Density\")\n",
    "\n",
    "ax.hist(\n",
    "    nbinom_dist.rvs(size=10000),\n",
    "    bins=bins,\n",
    "    density=True,\n",
    "    alpha=0.5,\n",
    "    rwidth=0.25,\n",
    "    align=\"mid\",\n",
    "    edgecolor=\"orange\",\n",
    "    color=\"orange\",\n",
    ")\n",
    "\n",
    "ax.plot(_x, nbinom_dist.pmf(_x), color=\"orange\", alpha=0.5)\n",
    "pass"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2e11220e",
   "metadata": {},
   "source": [
    "The fit isn't *perfect*, and with our relatively small sample size we wouldn't expect it to be. But it appears to represent our data fairly well."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "772fb9c2",
   "metadata": {},
   "source": [
    "# Using the negative binomial distribution for a Monte Carlo simulation\n",
    "\n",
    "One motivation for modeling the data with a negative binomial distribution is so that we can use it to generate a set of estimates using the \"Monte Carlo\" method — an approach similar to bootstrapping. Unlike the bootstrap method above, where we randomly select per capita filing rates from our pool of observations, we now randomly select them from our negative binomial distribution."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "524144d0",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:31.940711Z",
     "iopub.status.busy": "2021-09-15T18:32:31.937775Z",
     "iopub.status.idle": "2021-09-15T18:32:31.943011Z",
     "shell.execute_reply": "2021-09-15T18:32:31.943567Z"
    }
   },
   "outputs": [],
   "source": [
    "def gen_random_estimate_mc(populations, dist):\n",
    "    rand_rates = dist.rvs(size=len(populations))\n",
    "    return (populations * rand_rates).sum()\n",
    "\n",
    "def monte_carlo(populations, dist, n=1e4):\n",
    "    return pd.Series(gen_random_estimate_mc(populations, dist) for _ in range(int(n)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "f900049f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:31.951590Z",
     "iopub.status.busy": "2021-09-15T18:32:31.949304Z",
     "iopub.status.idle": "2021-09-15T18:32:35.169319Z",
     "shell.execute_reply": "2021-09-15T18:32:35.170125Z"
    }
   },
   "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>Bootstrap</th>\n",
       "      <th>Monte Carlo</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>10000</td>\n",
       "      <td>10000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>173618</td>\n",
       "      <td>173317</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>19294</td>\n",
       "      <td>19754</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>127230</td>\n",
       "      <td>114994</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.5%</th>\n",
       "      <td>142919</td>\n",
       "      <td>138835</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>158982</td>\n",
       "      <td>159509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>171300</td>\n",
       "      <td>171885</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>186030</td>\n",
       "      <td>185492</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97.5%</th>\n",
       "      <td>216686</td>\n",
       "      <td>216515</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>249941</td>\n",
       "      <td>271559</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Bootstrap  Monte Carlo\n",
       "count      10000        10000\n",
       "mean      173618       173317\n",
       "std        19294        19754\n",
       "min       127230       114994\n",
       "2.5%      142919       138835\n",
       "25%       158982       159509\n",
       "50%       171300       171885\n",
       "75%       186030       185492\n",
       "97.5%     216686       216515\n",
       "max       249941       271559"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mc_nbinom_not_counted = monte_carlo(\n",
    "    not_counted[\"pop_adult\"] / 1e5,\n",
    "    nbinom_dist,\n",
    "    n = 1e4\n",
    ")\n",
    "\n",
    "pd.DataFrame({\n",
    "    \"Bootstrap\": describe(bootstrapped_not_counted + known_total),\n",
    "    \"Monte Carlo\": describe(mc_nbinom_not_counted + known_total),\n",
    "})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bdefae58",
   "metadata": {},
   "source": [
    "Although the assumptions here are slightly different than for the bootstrap, and allow for a greater range of rates, the resulting estimates end up remarkable similar. Here are histograms of the two results, with the bootstrap in blue and the Monte Carlo simulation in orange:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "9d27d252",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:35.177392Z",
     "iopub.status.busy": "2021-09-15T18:32:35.176450Z",
     "iopub.status.idle": "2021-09-15T18:32:35.435934Z",
     "shell.execute_reply": "2021-09-15T18:32:35.435486Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "hist(bootstrapped_not_counted + known_total)\n",
    "hist(mc_nbinom_not_counted + known_total);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ea1884ed",
   "metadata": {},
   "source": [
    "# Incorporating data from states that do not disambiguate adults/minors\n",
    "\n",
    "Although we don't have fully-usable data from states that do not disambiguate between guardianships of adults and those of minors, we can still use this information to construct some bounds. There cannot be, for instance, more adult guardianship filings in these states than there are total adult+minor filings combined. And adult guardianship filings are likely to be some substantial portion of those filings. Although that portion is unknown, in states that do disambiguate those two types of cases the adult proportion appears to typically be between 50% and 95% of the total. Below, we generate a series of estimates that combine the following totals:\n",
    "\n",
    "- The known totals in our 23 disambiguated jurisdictions\n",
    "- The `Adults+Minors` totals for the states that provide them, multiplied by 0.5, 0.6, 0.7, 0.8, or 0.9\n",
    "- The results of a Monte Carlo simulation, using the same negative binomial distribution as above, for the remaining states\n",
    "\n",
    "This gives us a sense of the potential outcomes, given a range of adult:total ratios for the `Adults+Minors` states. Because we are multiplying all states by the same adult:total ratio, the distributions of each set of results is artificially narrow, but they provide a further sense of the possible national totals."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "efc85d4c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:35.452481Z",
     "iopub.status.busy": "2021-09-15T18:32:35.446050Z",
     "iopub.status.idle": "2021-09-15T18:32:35.455491Z",
     "shell.execute_reply": "2021-09-15T18:32:35.455914Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "12 states with Adults+Minors counts\n"
     ]
    },
    {
     "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>state</th>\n",
       "      <th>year</th>\n",
       "      <th>total</th>\n",
       "      <th>pop_adult</th>\n",
       "      <th>per_100k</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Alaska</td>\n",
       "      <td>2019</td>\n",
       "      <td>1084</td>\n",
       "      <td>551562</td>\n",
       "      <td>196.532756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>California</td>\n",
       "      <td>2019</td>\n",
       "      <td>16634</td>\n",
       "      <td>30617582</td>\n",
       "      <td>54.328261</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Delaware</td>\n",
       "      <td>2019</td>\n",
       "      <td>1744</td>\n",
       "      <td>770192</td>\n",
       "      <td>226.437044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Florida</td>\n",
       "      <td>2019</td>\n",
       "      <td>7906</td>\n",
       "      <td>17247808</td>\n",
       "      <td>45.837709</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Hawaii</td>\n",
       "      <td>2019</td>\n",
       "      <td>613</td>\n",
       "      <td>1116004</td>\n",
       "      <td>54.928119</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Indiana</td>\n",
       "      <td>2019</td>\n",
       "      <td>8301</td>\n",
       "      <td>5164245</td>\n",
       "      <td>160.739856</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Montana</td>\n",
       "      <td>2019</td>\n",
       "      <td>1129</td>\n",
       "      <td>840190</td>\n",
       "      <td>134.374368</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Nevada</td>\n",
       "      <td>2019</td>\n",
       "      <td>2513</td>\n",
       "      <td>2387517</td>\n",
       "      <td>105.255795</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>New York</td>\n",
       "      <td>2019</td>\n",
       "      <td>26763</td>\n",
       "      <td>15425262</td>\n",
       "      <td>173.501105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>North Carolina</td>\n",
       "      <td>2019</td>\n",
       "      <td>4649</td>\n",
       "      <td>8187369</td>\n",
       "      <td>56.782588</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Tennessee</td>\n",
       "      <td>2019</td>\n",
       "      <td>3754</td>\n",
       "      <td>5319123</td>\n",
       "      <td>70.575544</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Wisconsin</td>\n",
       "      <td>2019</td>\n",
       "      <td>6051</td>\n",
       "      <td>4555837</td>\n",
       "      <td>132.818624</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             state  year  total  pop_adult    per_100k\n",
       "0           Alaska  2019   1084     551562  196.532756\n",
       "1       California  2019  16634   30617582   54.328261\n",
       "2         Delaware  2019   1744     770192  226.437044\n",
       "3          Florida  2019   7906   17247808   45.837709\n",
       "4           Hawaii  2019    613    1116004   54.928119\n",
       "5          Indiana  2019   8301    5164245  160.739856\n",
       "6          Montana  2019   1129     840190  134.374368\n",
       "7           Nevada  2019   2513    2387517  105.255795\n",
       "8         New York  2019  26763   15425262  173.501105\n",
       "9   North Carolina  2019   4649    8187369   56.782588\n",
       "10       Tennessee  2019   3754    5319123   70.575544\n",
       "11       Wisconsin  2019   6051    4555837  132.818624"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "latest_totals_all_ages = get_totals_and_rates(\n",
    "    counts\n",
    "    .loc[lambda df: df[\"year\"] == 2019]\n",
    "    .loc[lambda df: df[\"for_total\"] == \"Y\"]\n",
    "    .loc[lambda df: df[\"age_group\"] == \"Adults+Minors\"]\n",
    ")\n",
    "\n",
    "assert latest_totals_all_ages[\"state\"].isin(latest_totals[\"state\"]).sum() == 0\n",
    "\n",
    "print(f\"{len(latest_totals_all_ages)} states with Adults+Minors counts\")\n",
    "latest_totals_all_ages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "8c2d933d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:35.466896Z",
     "iopub.status.busy": "2021-09-15T18:32:35.461345Z",
     "iopub.status.idle": "2021-09-15T18:32:39.607026Z",
     "shell.execute_reply": "2021-09-15T18:32:39.607952Z"
    }
   },
   "outputs": [],
   "source": [
    "mc_nbinom_no_data = monte_carlo(\n",
    "    (\n",
    "        populations\n",
    "        .loc[lambda df: ~df[\"state\"].isin(latest_totals[\"state\"])]\n",
    "        .loc[lambda df: ~df[\"state\"].isin(latest_totals_all_ages[\"state\"])]\n",
    "        [\"pop_adult\"]\n",
    "    ) / 1e5,\n",
    "    nbinom_dist,\n",
    "    n = 1e4\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "ea567a35",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:39.616185Z",
     "iopub.status.busy": "2021-09-15T18:32:39.614339Z",
     "iopub.status.idle": "2021-09-15T18:32:39.641317Z",
     "shell.execute_reply": "2021-09-15T18:32:39.642112Z"
    }
   },
   "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>50%</th>\n",
       "      <th>60%</th>\n",
       "      <th>70%</th>\n",
       "      <th>80%</th>\n",
       "      <th>90%</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>10000</td>\n",
       "      <td>10000</td>\n",
       "      <td>10000</td>\n",
       "      <td>10000</td>\n",
       "      <td>10000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>150645</td>\n",
       "      <td>158759</td>\n",
       "      <td>166873</td>\n",
       "      <td>174987</td>\n",
       "      <td>183101</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>11030</td>\n",
       "      <td>11030</td>\n",
       "      <td>11030</td>\n",
       "      <td>11030</td>\n",
       "      <td>11030</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>120251</td>\n",
       "      <td>128365</td>\n",
       "      <td>136479</td>\n",
       "      <td>144593</td>\n",
       "      <td>152707</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.5%</th>\n",
       "      <td>132219</td>\n",
       "      <td>140333</td>\n",
       "      <td>148447</td>\n",
       "      <td>156561</td>\n",
       "      <td>164675</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>142858</td>\n",
       "      <td>150972</td>\n",
       "      <td>159086</td>\n",
       "      <td>167200</td>\n",
       "      <td>175315</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>149594</td>\n",
       "      <td>157708</td>\n",
       "      <td>165822</td>\n",
       "      <td>173936</td>\n",
       "      <td>182050</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>157073</td>\n",
       "      <td>165187</td>\n",
       "      <td>173301</td>\n",
       "      <td>181415</td>\n",
       "      <td>189529</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97.5%</th>\n",
       "      <td>175776</td>\n",
       "      <td>183890</td>\n",
       "      <td>192004</td>\n",
       "      <td>200119</td>\n",
       "      <td>208233</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>205186</td>\n",
       "      <td>213301</td>\n",
       "      <td>221415</td>\n",
       "      <td>229529</td>\n",
       "      <td>237643</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          50%     60%     70%     80%     90%\n",
       "count   10000   10000   10000   10000   10000\n",
       "mean   150645  158759  166873  174987  183101\n",
       "std     11030   11030   11030   11030   11030\n",
       "min    120251  128365  136479  144593  152707\n",
       "2.5%   132219  140333  148447  156561  164675\n",
       "25%    142858  150972  159086  167200  175315\n",
       "50%    149594  157708  165822  173936  182050\n",
       "75%    157073  165187  173301  181415  189529\n",
       "97.5%  175776  183890  192004  200119  208233\n",
       "max    205186  213301  221415  229529  237643"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame({\n",
    "    f\"{x*100:.0f}%\": describe(mc_nbinom_no_data + known_total + latest_totals_all_ages[\"total\"].sum() * x) \n",
    "    for x in np.arange(0.5, 1, 0.1)\n",
    "})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f3ded250",
   "metadata": {},
   "source": [
    "# Examining prior years\n",
    "\n",
    "Although our data is most complete from 2019, which is also the most recent pre-pandemic year, we might be intersted to see how the rates have has changed from prior years. To do so, we calculate the negative binomial distribution for 2017 and 2018 and compare it to 2019. Such comparisons should be taken lightly, since the sample is different for each year, but they seem to indicate that filing rates have likely increased during this time: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "27f6dd63",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:39.649359Z",
     "iopub.status.busy": "2021-09-15T18:32:39.644682Z",
     "iopub.status.idle": "2021-09-15T18:32:39.684639Z",
     "shell.execute_reply": "2021-09-15T18:32:39.685170Z"
    }
   },
   "outputs": [],
   "source": [
    "nbinom_dist_years = pd.Series({ year:(\n",
    "    nbinom_from_rates(\n",
    "        get_totals_and_rates(\n",
    "            counts\n",
    "            .loc[lambda df: df[\"year\"] == year]\n",
    "            .loc[lambda df: df[\"for_total\"] == \"Y\"]\n",
    "            .loc[lambda df: df[\"age_group\"] == \"Adults ONLY\"]\n",
    "        )\n",
    "        [\"per_100k\"]\n",
    "    )\n",
    ") for year in [ 2017, 2018, 2019 ]})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "621fcc66",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:39.694171Z",
     "iopub.status.busy": "2021-09-15T18:32:39.693213Z",
     "iopub.status.idle": "2021-09-15T18:32:39.701855Z",
     "shell.execute_reply": "2021-09-15T18:32:39.702292Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2017    57.9\n",
       "2018    63.7\n",
       "2019    69.0\n",
       "dtype: float64"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nbinom_dist_years.apply(lambda dist: dist.mean()).round(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "067b2d13",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:39.710459Z",
     "iopub.status.busy": "2021-09-15T18:32:39.704769Z",
     "iopub.status.idle": "2021-09-15T18:32:40.014653Z",
     "shell.execute_reply": "2021-09-15T18:32:40.015177Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "nbinom_pmf_years = nbinom_dist_years.apply(lambda dist: pd.Series(dist.pmf(_x))).T\n",
    "nbinom_pmf_years.loc[:,:2019].plot();"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d5dd283f",
   "metadata": {},
   "source": [
    "Another, more direct way of looking at this is to compare filing rates in both 2018 and 2019 for the jurisdictions where we have both counts. Here, we see a slight increase in the majority of jurisdictions and overall:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "d6a88591",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:40.028138Z",
     "iopub.status.busy": "2021-09-15T18:32:40.017613Z",
     "iopub.status.idle": "2021-09-15T18:32:40.071707Z",
     "shell.execute_reply": "2021-09-15T18:32:40.072346Z"
    }
   },
   "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>year</th>\n",
       "      <th>2018</th>\n",
       "      <th>2019</th>\n",
       "      <th>change</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>state</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Kentucky</th>\n",
       "      <td>35.62</td>\n",
       "      <td>43.09</td>\n",
       "      <td>7.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Nebraska</th>\n",
       "      <td>138.31</td>\n",
       "      <td>144.62</td>\n",
       "      <td>6.31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Vermont</th>\n",
       "      <td>96.67</td>\n",
       "      <td>100.40</td>\n",
       "      <td>3.73</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Utah</th>\n",
       "      <td>32.49</td>\n",
       "      <td>36.18</td>\n",
       "      <td>3.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Washington</th>\n",
       "      <td>47.06</td>\n",
       "      <td>50.19</td>\n",
       "      <td>3.13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Kansas</th>\n",
       "      <td>49.07</td>\n",
       "      <td>51.60</td>\n",
       "      <td>2.53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ohio</th>\n",
       "      <td>74.22</td>\n",
       "      <td>76.73</td>\n",
       "      <td>2.51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>New Mexico</th>\n",
       "      <td>40.96</td>\n",
       "      <td>42.94</td>\n",
       "      <td>1.97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Michigan</th>\n",
       "      <td>159.77</td>\n",
       "      <td>161.36</td>\n",
       "      <td>1.58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Idaho</th>\n",
       "      <td>43.47</td>\n",
       "      <td>44.37</td>\n",
       "      <td>0.90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>District of Columbia</th>\n",
       "      <td>98.86</td>\n",
       "      <td>99.55</td>\n",
       "      <td>0.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Minnesota</th>\n",
       "      <td>60.63</td>\n",
       "      <td>61.29</td>\n",
       "      <td>0.67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Virginia</th>\n",
       "      <td>37.13</td>\n",
       "      <td>37.72</td>\n",
       "      <td>0.60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Massachusetts</th>\n",
       "      <td>72.71</td>\n",
       "      <td>73.27</td>\n",
       "      <td>0.56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Oregon</th>\n",
       "      <td>39.93</td>\n",
       "      <td>40.43</td>\n",
       "      <td>0.51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Arkansas</th>\n",
       "      <td>49.75</td>\n",
       "      <td>50.22</td>\n",
       "      <td>0.47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Colorado</th>\n",
       "      <td>31.16</td>\n",
       "      <td>31.49</td>\n",
       "      <td>0.33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Pennsylvania</th>\n",
       "      <td>33.59</td>\n",
       "      <td>32.50</td>\n",
       "      <td>-1.09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>New Hampshire</th>\n",
       "      <td>86.83</td>\n",
       "      <td>85.56</td>\n",
       "      <td>-1.27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Missouri</th>\n",
       "      <td>64.55</td>\n",
       "      <td>61.30</td>\n",
       "      <td>-3.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>North Dakota</th>\n",
       "      <td>59.12</td>\n",
       "      <td>53.96</td>\n",
       "      <td>-5.16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Arizona</th>\n",
       "      <td>49.66</td>\n",
       "      <td>42.26</td>\n",
       "      <td>-7.40</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "year                    2018    2019  change\n",
       "state                                       \n",
       "Kentucky               35.62   43.09    7.48\n",
       "Nebraska              138.31  144.62    6.31\n",
       "Vermont                96.67  100.40    3.73\n",
       "Utah                   32.49   36.18    3.69\n",
       "Washington             47.06   50.19    3.13\n",
       "Kansas                 49.07   51.60    2.53\n",
       "Ohio                   74.22   76.73    2.51\n",
       "New Mexico             40.96   42.94    1.97\n",
       "Michigan              159.77  161.36    1.58\n",
       "Idaho                  43.47   44.37    0.90\n",
       "District of Columbia   98.86   99.55    0.69\n",
       "Minnesota              60.63   61.29    0.67\n",
       "Virginia               37.13   37.72    0.60\n",
       "Massachusetts          72.71   73.27    0.56\n",
       "Oregon                 39.93   40.43    0.51\n",
       "Arkansas               49.75   50.22    0.47\n",
       "Colorado               31.16   31.49    0.33\n",
       "Pennsylvania           33.59   32.50   -1.09\n",
       "New Hampshire          86.83   85.56   -1.27\n",
       "Missouri               64.55   61.30   -3.25\n",
       "North Dakota           59.12   53.96   -5.16\n",
       "Arizona                49.66   42.26   -7.40"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "change_2018_19 = (\n",
    "    get_totals_and_rates(\n",
    "        counts\n",
    "        .loc[lambda df: df[\"for_total\"] == \"Y\"]\n",
    "        .loc[lambda df: df[\"age_group\"] == \"Adults ONLY\"]        \n",
    "    )\n",
    "    .loc[lambda df: df[\"year\"].isin([ 2018, 2019 ])]\n",
    "    .set_index([\n",
    "        \"state\",\n",
    "        \"year\"\n",
    "    ])\n",
    "    [\"per_100k\"]\n",
    "    .unstack()\n",
    "    .dropna()\n",
    "    .assign(\n",
    "        change = lambda df: df[2019] - df[2018]\n",
    "    )\n",
    ")\n",
    "\n",
    "change_2018_19.sort_values(\"change\", ascending=False).round(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "0adc7616",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:40.082623Z",
     "iopub.status.busy": "2021-09-15T18:32:40.081520Z",
     "iopub.status.idle": "2021-09-15T18:32:40.088990Z",
     "shell.execute_reply": "2021-09-15T18:32:40.089635Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    22.000000\n",
       "mean      0.886168\n",
       "std       3.326963\n",
       "min      -7.395609\n",
       "25%       0.368698\n",
       "50%       0.680645\n",
       "75%       2.526177\n",
       "max       7.475175\n",
       "Name: change, dtype: float64"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "change_2018_19[\"change\"].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "734f5d7e",
   "metadata": {},
   "source": [
    "# Appendix A: Adjusting for population size\n",
    "\n",
    "Our sample of jurisdictions for which we have fully usable data spans a broad range of populations, but slightly overrepresents smaller states and slightly underrepresents larger states. We can this see by grouping all jurisdictions into three population terciles and comparing the size-distribution in and out of our sample:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "3567e92e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:40.098596Z",
     "iopub.status.busy": "2021-09-15T18:32:40.097399Z",
     "iopub.status.idle": "2021-09-15T18:32:40.138887Z",
     "shell.execute_reply": "2021-09-15T18:32:40.139510Z"
    }
   },
   "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>in_sample</th>\n",
       "      <th>False</th>\n",
       "      <th>True</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>(445024.999, 2254204.0]</th>\n",
       "      <td>0.32</td>\n",
       "      <td>0.35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>(2254204.0, 4899310.333]</th>\n",
       "      <td>0.32</td>\n",
       "      <td>0.35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>(4899310.333, 30617582.0]</th>\n",
       "      <td>0.36</td>\n",
       "      <td>0.30</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "in_sample                  False  True\n",
       "(445024.999, 2254204.0]     0.32  0.35\n",
       "(2254204.0, 4899310.333]    0.32  0.35\n",
       "(4899310.333, 30617582.0]   0.36  0.30"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    populations\n",
    "    .assign(\n",
    "        in_sample = lambda df: df[\"state\"].isin(latest_totals[\"state\"]),\n",
    "        pop_adult_tercile = lambda df: pd.qcut(df[\"pop_adult\"], 3),\n",
    "    )\n",
    "    .groupby(\"in_sample\")\n",
    "    [\"pop_adult_tercile\"]\n",
    "    .value_counts(normalize=True)\n",
    "    .unstack()\n",
    "    .T\n",
    "    .round(2)\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bc9f2bab",
   "metadata": {},
   "source": [
    "We also observe a *potential* negative correlation between population size and filings per capita:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "1c491e95",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:40.150382Z",
     "iopub.status.busy": "2021-09-15T18:32:40.149391Z",
     "iopub.status.idle": "2021-09-15T18:32:40.562018Z",
     "shell.execute_reply": "2021-09-15T18:32:40.562656Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "(\n",
    "    latest_totals\n",
    "    .pipe(lambda df: sns.regplot(\n",
    "        x = df[\"pop_adult\"].apply(np.log10),\n",
    "        y = df[\"per_100k\"],\n",
    "        truncate = False,\n",
    "        scatter_kws = { \"alpha\": 0.5 }\n",
    "    ))\n",
    ")\n",
    "pass"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2048f6a0",
   "metadata": {},
   "source": [
    "The relationship does not reach standard levels of statistical significance, although our observations may be underpowered (i.e., sample size too small) to reach that level:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "71cb048b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:40.577493Z",
     "iopub.status.busy": "2021-09-15T18:32:40.576784Z",
     "iopub.status.idle": "2021-09-15T18:32:40.605349Z",
     "shell.execute_reply": "2021-09-15T18:32:40.605810Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<caption>OLS Regression Results</caption>\n",
       "<tr>\n",
       "  <th>Dep. Variable:</th>        <td>per_100k</td>     <th>  R-squared:         </th> <td>   0.028</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model:</th>                   <td>OLS</td>       <th>  Adj. R-squared:    </th> <td>  -0.019</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Method:</th>             <td>Least Squares</td>  <th>  F-statistic:       </th> <td>  0.5992</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Date:</th>             <td>Fri, 17 Sep 2021</td> <th>  Prob (F-statistic):</th>  <td> 0.447</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Time:</th>                 <td>10:23:20</td>     <th>  Log-Likelihood:    </th> <td> -116.74</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>No. Observations:</th>      <td>    23</td>      <th>  AIC:               </th> <td>   237.5</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Df Residuals:</th>          <td>    21</td>      <th>  BIC:               </th> <td>   239.8</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Df Model:</th>              <td>     1</td>      <th>                     </th>     <td> </td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Covariance Type:</th>      <td>nonrobust</td>    <th>                     </th>     <td> </td>   \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "      <td></td>         <th>coef</th>     <th>std err</th>      <th>t</th>      <th>P>|t|</th>  <th>[0.025</th>    <th>0.975]</th>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>const</th>     <td>  180.9806</td> <td>  144.923</td> <td>    1.249</td> <td> 0.225</td> <td> -120.403</td> <td>  482.364</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>pop_adult</th> <td>  -17.3712</td> <td>   22.440</td> <td>   -0.774</td> <td> 0.447</td> <td>  -64.038</td> <td>   29.296</td>\n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "  <th>Omnibus:</th>       <td>12.103</td> <th>  Durbin-Watson:     </th> <td>   1.905</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(Omnibus):</th> <td> 0.002</td> <th>  Jarque-Bera (JB):  </th> <td>  10.388</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Skew:</th>          <td> 1.520</td> <th>  Prob(JB):          </th> <td> 0.00555</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Kurtosis:</th>      <td> 4.263</td> <th>  Cond. No.          </th> <td>    113.</td>\n",
       "</tr>\n",
       "</table><br/><br/>Notes:<br/>[1] Standard Errors assume that the covariance matrix of the errors is correctly specified."
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary.Summary'>\n",
       "\"\"\"\n",
       "                            OLS Regression Results                            \n",
       "==============================================================================\n",
       "Dep. Variable:               per_100k   R-squared:                       0.028\n",
       "Model:                            OLS   Adj. R-squared:                 -0.019\n",
       "Method:                 Least Squares   F-statistic:                    0.5992\n",
       "Date:                Fri, 17 Sep 2021   Prob (F-statistic):              0.447\n",
       "Time:                        10:23:20   Log-Likelihood:                -116.74\n",
       "No. Observations:                  23   AIC:                             237.5\n",
       "Df Residuals:                      21   BIC:                             239.8\n",
       "Df Model:                           1                                         \n",
       "Covariance Type:            nonrobust                                         \n",
       "==============================================================================\n",
       "                 coef    std err          t      P>|t|      [0.025      0.975]\n",
       "------------------------------------------------------------------------------\n",
       "const        180.9806    144.923      1.249      0.225    -120.403     482.364\n",
       "pop_adult    -17.3712     22.440     -0.774      0.447     -64.038      29.296\n",
       "==============================================================================\n",
       "Omnibus:                       12.103   Durbin-Watson:                   1.905\n",
       "Prob(Omnibus):                  0.002   Jarque-Bera (JB):               10.388\n",
       "Skew:                           1.520   Prob(JB):                      0.00555\n",
       "Kurtosis:                       4.263   Cond. No.                         113.\n",
       "==============================================================================\n",
       "\n",
       "Notes:\n",
       "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
       "\"\"\""
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pop_ols = (\n",
    "    latest_totals\n",
    "    .pipe(lambda df: (\n",
    "        sm.OLS(\n",
    "            df[\"per_100k\"],\n",
    "            sm.add_constant(df[\"pop_adult\"].apply(np.log10)),\n",
    "        )\n",
    "     ))\n",
    "    .fit()\n",
    ")\n",
    "\n",
    "pop_ols.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "57c3bdc8",
   "metadata": {},
   "source": [
    "It's unclear whether the relationship is causal, or whether it even holds across the full set of states. But to test the sensitivity our estimates to such a possible influence, we can repeat our bootstrap approach while adjusting for population size.\n",
    "\n",
    "We'll use the observed correlation above; it is somewhat arbitrary, but should serve as a useful demonstration.\n",
    "\n",
    "We calculate the adjustment in this manner: `coef * (log10(state's adult population) - log10(average state adult population))`, where `coef` is the coefficient for `log10(adult_pop)` in the regression above."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "dcd6cdfb",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:40.620969Z",
     "iopub.status.busy": "2021-09-15T18:32:40.620159Z",
     "iopub.status.idle": "2021-09-15T18:32:40.624891Z",
     "shell.execute_reply": "2021-09-15T18:32:40.623432Z"
    }
   },
   "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>state</th>\n",
       "      <th>pop_total</th>\n",
       "      <th>pop_adult</th>\n",
       "      <th>pop_adult_log10</th>\n",
       "      <th>adj</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Alabama</td>\n",
       "      <td>4903185</td>\n",
       "      <td>3814879</td>\n",
       "      <td>6.581481</td>\n",
       "      <td>-1.756828</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Alaska</td>\n",
       "      <td>731545</td>\n",
       "      <td>551562</td>\n",
       "      <td>5.741594</td>\n",
       "      <td>12.832977</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Arizona</td>\n",
       "      <td>7278717</td>\n",
       "      <td>5638481</td>\n",
       "      <td>6.751162</td>\n",
       "      <td>-4.704390</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Arkansas</td>\n",
       "      <td>3017804</td>\n",
       "      <td>2317649</td>\n",
       "      <td>6.365048</td>\n",
       "      <td>2.002867</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>California</td>\n",
       "      <td>39512223</td>\n",
       "      <td>30617582</td>\n",
       "      <td>7.485971</td>\n",
       "      <td>-17.468873</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        state  pop_total  pop_adult  pop_adult_log10        adj\n",
       "0     Alabama    4903185    3814879         6.581481  -1.756828\n",
       "1      Alaska     731545     551562         5.741594  12.832977\n",
       "2     Arizona    7278717    5638481         6.751162  -4.704390\n",
       "3    Arkansas    3017804    2317649         6.365048   2.002867\n",
       "4  California   39512223   30617582         7.485971 -17.468873"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "adjustment_table = (\n",
    "    populations\n",
    "    .assign(\n",
    "        pop_adult_log10 = lambda df: df[\"pop_adult\"].apply(np.log10),\n",
    "        adj = lambda df: (\n",
    "            pop_ols.params[\"pop_adult\"] * \n",
    "            (df[\"pop_adult_log10\"] - df[\"pop_adult_log10\"].mean())\n",
    "        )\n",
    "    )\n",
    ")\n",
    "\n",
    "adjustment_table.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "0176834e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:40.630552Z",
     "iopub.status.busy": "2021-09-15T18:32:40.629461Z",
     "iopub.status.idle": "2021-09-15T18:32:40.631850Z",
     "shell.execute_reply": "2021-09-15T18:32:40.632394Z"
    }
   },
   "outputs": [],
   "source": [
    "# Make sure that the positive and negative adjustments balance out\n",
    "assert int(adjustment_table[\"adj\"].sum()) == 0"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bbf43bfa",
   "metadata": {},
   "source": [
    "Next, we subtract the adjustment factor from each of our observed per capita filiing rates to generate a set of normalized rates, effectively asking \"What might this rate be if the state had the average population?\" \n",
    "\n",
    "We also plot the normalized rates (in orange) against the observed rates (in blue) to visualize the shift, and we observe that it acheives our goal of flattening the regression line:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "b3b3d659",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:40.669636Z",
     "iopub.status.busy": "2021-09-15T18:32:40.667948Z",
     "iopub.status.idle": "2021-09-15T18:32:41.111567Z",
     "shell.execute_reply": "2021-09-15T18:32:41.112101Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "latest_totals_adj = (\n",
    "    latest_totals\n",
    "    .merge(\n",
    "        adjustment_table\n",
    "        [[\n",
    "            \"state\",\n",
    "            \"pop_adult_log10\",\n",
    "            \"adj\"\n",
    "        ]],\n",
    "        how = \"left\"\n",
    "    )\n",
    "    .assign(\n",
    "        per_100k_adj = lambda df: df[\"per_100k\"] - df[\"adj\"]\n",
    "    )\n",
    ")\n",
    "\n",
    "(\n",
    "    latest_totals_adj\n",
    "    .pipe(lambda df: sns.regplot(\n",
    "        y = df[\"per_100k\"],\n",
    "        x = df[\"pop_adult_log10\"],\n",
    "        truncate = False,\n",
    "        scatter_kws = { \"alpha\": 0.25 }\n",
    "    ))\n",
    ")\n",
    "\n",
    "(\n",
    "    latest_totals_adj\n",
    "    .pipe(lambda df: sns.regplot(\n",
    "        y = df[\"per_100k_adj\"],\n",
    "        x = df[\"pop_adult_log10\"],\n",
    "        truncate = False,\n",
    "        scatter_kws = { \"alpha\": 0.5 }\n",
    "    ))\n",
    ")\n",
    "\n",
    "pass"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "56598350",
   "metadata": {},
   "source": [
    "Finally, we modify the bootstrap approach to take these adjustments into account. Specifically, for each sub-iteration, we pair a state with a random draw of one of the *normalized* rates, but then add the adjustment (increasing the effective rate for small states and decreasing it for large states) before calculating the product."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "d1eff566",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:41.117599Z",
     "iopub.status.busy": "2021-09-15T18:32:41.116855Z",
     "iopub.status.idle": "2021-09-15T18:32:41.118959Z",
     "shell.execute_reply": "2021-09-15T18:32:41.119448Z"
    }
   },
   "outputs": [],
   "source": [
    "def gen_random_estimate_with_adj(populations, adj, rates):\n",
    "    rand_rates = gen_random_items(rates, len(populations))\n",
    "    return ((populations * (rand_rates + adj))).sum()\n",
    "\n",
    "def bootstrap_with_adj(populations, adj, rates, n=1e4):\n",
    "    return pd.Series(\n",
    "        gen_random_estimate_with_adj(populations, adj, rates)\n",
    "        for _ in range(int(n))\n",
    "    )"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3a242cad",
   "metadata": {},
   "source": [
    "This produces, as one would expect given the negative coefficient and the under-representation of large states in our sample, somewhat smaller estimates. Although, as noted above, unclear whether adjusting for population sizes produces a more accurate estimate, these results help provide an understanding of how the estimates might change *if* there were an underlying mechanism that decreased filing rates in larger states:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "af39e6e7",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:41.125002Z",
     "iopub.status.busy": "2021-09-15T18:32:41.123937Z",
     "iopub.status.idle": "2021-09-15T18:32:41.126634Z",
     "shell.execute_reply": "2021-09-15T18:32:41.127213Z"
    }
   },
   "outputs": [],
   "source": [
    "adjustment_table_not_counted = (\n",
    "    adjustment_table\n",
    "    .loc[lambda df: df[\"state\"].isin(not_counted[\"state\"])]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "267d8c61",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:41.170531Z",
     "iopub.status.busy": "2021-09-15T18:32:41.169611Z",
     "iopub.status.idle": "2021-09-15T18:32:46.598973Z",
     "shell.execute_reply": "2021-09-15T18:32:46.598386Z"
    }
   },
   "outputs": [],
   "source": [
    "bootstrapped_adj_not_counted = bootstrap_with_adj(\n",
    "    adjustment_table_not_counted[\"pop_adult\"] / 1e5,\n",
    "    adjustment_table_not_counted[\"adj\"],\n",
    "    latest_totals_adj[\"per_100k_adj\"],\n",
    "    1e4\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "349e042d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:46.604016Z",
     "iopub.status.busy": "2021-09-15T18:32:46.603388Z",
     "iopub.status.idle": "2021-09-15T18:32:46.619103Z",
     "shell.execute_reply": "2021-09-15T18:32:46.619610Z"
    }
   },
   "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>Bootstrap</th>\n",
       "      <th>Monte Carlo</th>\n",
       "      <th>Pop.-adj. Bootstrap</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>10000</td>\n",
       "      <td>10000</td>\n",
       "      <td>10000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>173618</td>\n",
       "      <td>173317</td>\n",
       "      <td>156942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>19294</td>\n",
       "      <td>19754</td>\n",
       "      <td>18724</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>127230</td>\n",
       "      <td>114994</td>\n",
       "      <td>115434</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2.5%</th>\n",
       "      <td>142919</td>\n",
       "      <td>138835</td>\n",
       "      <td>127557</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>158982</td>\n",
       "      <td>159509</td>\n",
       "      <td>143092</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>171300</td>\n",
       "      <td>171885</td>\n",
       "      <td>154506</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>186030</td>\n",
       "      <td>185492</td>\n",
       "      <td>168509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97.5%</th>\n",
       "      <td>216686</td>\n",
       "      <td>216515</td>\n",
       "      <td>199750</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>249941</td>\n",
       "      <td>271559</td>\n",
       "      <td>238711</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Bootstrap  Monte Carlo  Pop.-adj. Bootstrap\n",
       "count      10000        10000                10000\n",
       "mean      173618       173317               156942\n",
       "std        19294        19754                18724\n",
       "min       127230       114994               115434\n",
       "2.5%      142919       138835               127557\n",
       "25%       158982       159509               143092\n",
       "50%       171300       171885               154506\n",
       "75%       186030       185492               168509\n",
       "97.5%     216686       216515               199750\n",
       "max       249941       271559               238711"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame({\n",
    "    \"Bootstrap\": describe(bootstrapped_not_counted + known_total),\n",
    "    \"Monte Carlo\": describe(mc_nbinom_not_counted + known_total),\n",
    "    \"Pop.-adj. Bootstrap\": describe(bootstrapped_adj_not_counted + known_total),\n",
    "})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5792943a",
   "metadata": {},
   "source": [
    "Here we compare the initial bootstrap estimates (in blue) to the population-adjusted ones (in orange):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "8b0fac77",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:46.626345Z",
     "iopub.status.busy": "2021-09-15T18:32:46.625611Z",
     "iopub.status.idle": "2021-09-15T18:32:46.869804Z",
     "shell.execute_reply": "2021-09-15T18:32:46.870280Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "hist(bootstrapped_not_counted + known_total)\n",
    "hist(bootstrapped_adj_not_counted + known_total)\n",
    "pass"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c0f491eb",
   "metadata": {},
   "source": [
    "# Appendix B: Incorporating information about legal requirements\n",
    "\n",
    "Various factors likely influence the number of guardianship cases filed in a given state. Some state laws make filing guardianship cases harder than others, for instance by requiring certain types of documentation. Others try to create alternatives to guardianship, such as through [Supported Decision-Making Agreements](https://www.americanbar.org/groups/law_aging/publications/bifocal/vol-41/volume-41-issue-1/where-states-stand-on-supported-decision-making/). BuzzFeed News collected data on four such aspects of guardianship law for each jurisdiction, and compared them to the observed filing rates:\n",
    "\n",
    "- `sdma`: Did the state allow for Supported Decision-Making Agreements in 2019?\n",
    "- `lra`: Did the state require judges to consider less restrictive alternatives to guardianship before granting a petition in 2019?\n",
    "- `ex_parte`: Did the state allow for ex parte orders of guardianship in 2019?\n",
    "- `documentation_required`: Did the state require documentation of incapacity to be submitted with the guardianship petition in 2019?\n",
    "\n",
    "Although it would be reasonable to assume these laws have some effect, we saw no clear pattern we could confidently incorporate into our estimates, perhaps due to our small sample size or perhaps because other factors outweigh these."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "9b99aefd",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:46.879949Z",
     "iopub.status.busy": "2021-09-15T18:32:46.873432Z",
     "iopub.status.idle": "2021-09-15T18:32:47.037379Z",
     "shell.execute_reply": "2021-09-15T18:32:47.037882Z"
    }
   },
   "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>state</th>\n",
       "      <th>sdma</th>\n",
       "      <th>lra</th>\n",
       "      <th>ex_parte</th>\n",
       "      <th>docu_req</th>\n",
       "      <th>num_constraints</th>\n",
       "      <th>has_constraint</th>\n",
       "      <th>per_100k</th>\n",
       "      <th>pop_adult</th>\n",
       "      <th>pop_adult_log10</th>\n",
       "      <th>in_sample</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Connecticut</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>165.618513</td>\n",
       "      <td>2837847</td>\n",
       "      <td>6.452989</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>Michigan</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "      <td>161.355637</td>\n",
       "      <td>7842924</td>\n",
       "      <td>6.894478</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>Nebraska</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>144.617077</td>\n",
       "      <td>1458334</td>\n",
       "      <td>6.163857</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>Vermont</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "      <td>100.395307</td>\n",
       "      <td>509984</td>\n",
       "      <td>5.707557</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>District of Columbia</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>True</td>\n",
       "      <td>99.553136</td>\n",
       "      <td>577581</td>\n",
       "      <td>5.761613</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>New Hampshire</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>True</td>\n",
       "      <td>85.562330</td>\n",
       "      <td>1104458</td>\n",
       "      <td>6.043149</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>Ohio</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "      <td>76.730741</td>\n",
       "      <td>9111081</td>\n",
       "      <td>6.959570</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>Massachusetts</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>True</td>\n",
       "      <td>73.271076</td>\n",
       "      <td>5539703</td>\n",
       "      <td>6.743486</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>Missouri</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>True</td>\n",
       "      <td>61.298432</td>\n",
       "      <td>4766843</td>\n",
       "      <td>6.678231</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>Minnesota</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>61.294023</td>\n",
       "      <td>4336475</td>\n",
       "      <td>6.637137</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>North Dakota</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>True</td>\n",
       "      <td>53.961996</td>\n",
       "      <td>581891</td>\n",
       "      <td>5.764842</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>Kansas</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>51.602665</td>\n",
       "      <td>2213064</td>\n",
       "      <td>6.344994</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Arkansas</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>50.223308</td>\n",
       "      <td>2317649</td>\n",
       "      <td>6.365048</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>Washington</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>50.186228</td>\n",
       "      <td>5951832</td>\n",
       "      <td>6.774651</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Idaho</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "      <td>44.365970</td>\n",
       "      <td>1338864</td>\n",
       "      <td>6.126736</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>Kentucky</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>43.090485</td>\n",
       "      <td>3464802</td>\n",
       "      <td>6.539678</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>New Mexico</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>42.936697</td>\n",
       "      <td>1620991</td>\n",
       "      <td>6.209781</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Arizona</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>42.263156</td>\n",
       "      <td>5638481</td>\n",
       "      <td>6.751162</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>Oregon</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>40.433579</td>\n",
       "      <td>3351175</td>\n",
       "      <td>6.525197</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>Virginia</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>37.724706</td>\n",
       "      <td>6674671</td>\n",
       "      <td>6.824430</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>Utah</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>36.179418</td>\n",
       "      <td>2274774</td>\n",
       "      <td>6.356938</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>Pennsylvania</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>True</td>\n",
       "      <td>32.496093</td>\n",
       "      <td>10167376</td>\n",
       "      <td>7.007209</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Colorado</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>31.494369</td>\n",
       "      <td>4499217</td>\n",
       "      <td>6.653137</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   state  sdma  lra  ex_parte  docu_req  num_constraints  \\\n",
       "6            Connecticut     0    1         1         0                1   \n",
       "22              Michigan     0    0         1         0                0   \n",
       "27              Nebraska     0    1         1         0                1   \n",
       "45               Vermont     0    0         1         0                0   \n",
       "8   District of Columbia     1    1         1         0                2   \n",
       "29         New Hampshire     0    1         1         1                2   \n",
       "35                  Ohio     0    0         1         0                0   \n",
       "21         Massachusetts     0    1         1         1                2   \n",
       "25              Missouri     0    1         0         0                2   \n",
       "23             Minnesota     0    1         1         0                1   \n",
       "34          North Dakota     1    1         1         0                2   \n",
       "16                Kansas     0    1         1         0                1   \n",
       "3               Arkansas     0    1         1         0                1   \n",
       "47            Washington     0    1         1         0                1   \n",
       "12                 Idaho     0    0         1         0                0   \n",
       "17              Kentucky     0    0         0         0                1   \n",
       "31            New Mexico     0    1         1         0                1   \n",
       "2                Arizona     0    1         1         0                1   \n",
       "37                Oregon     0    1         1         0                1   \n",
       "46              Virginia     0    1         1         0                1   \n",
       "44                  Utah     0    1         1         0                1   \n",
       "38          Pennsylvania     0    1         1         1                2   \n",
       "5               Colorado     0    1         1         0                1   \n",
       "\n",
       "    has_constraint    per_100k  pop_adult  pop_adult_log10  in_sample  \n",
       "6             True  165.618513    2837847         6.452989       True  \n",
       "22           False  161.355637    7842924         6.894478       True  \n",
       "27            True  144.617077    1458334         6.163857       True  \n",
       "45           False  100.395307     509984         5.707557       True  \n",
       "8             True   99.553136     577581         5.761613       True  \n",
       "29            True   85.562330    1104458         6.043149       True  \n",
       "35           False   76.730741    9111081         6.959570       True  \n",
       "21            True   73.271076    5539703         6.743486       True  \n",
       "25            True   61.298432    4766843         6.678231       True  \n",
       "23            True   61.294023    4336475         6.637137       True  \n",
       "34            True   53.961996     581891         5.764842       True  \n",
       "16            True   51.602665    2213064         6.344994       True  \n",
       "3             True   50.223308    2317649         6.365048       True  \n",
       "47            True   50.186228    5951832         6.774651       True  \n",
       "12           False   44.365970    1338864         6.126736       True  \n",
       "17            True   43.090485    3464802         6.539678       True  \n",
       "31            True   42.936697    1620991         6.209781       True  \n",
       "2             True   42.263156    5638481         6.751162       True  \n",
       "37            True   40.433579    3351175         6.525197       True  \n",
       "46            True   37.724706    6674671         6.824430       True  \n",
       "44            True   36.179418    2274774         6.356938       True  \n",
       "38            True   32.496093   10167376         7.007209       True  \n",
       "5             True   31.494369    4499217         6.653137       True  "
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "regimes = (\n",
    "    pd.read_csv(\"../data/raw/legal-regimes.csv\")\n",
    "    [[\n",
    "        \"state\",\n",
    "        \"sdma\",\n",
    "        \"lra\",\n",
    "        \"ex_parte\",\n",
    "        \"documentation_required\",\n",
    "    ]]\n",
    "    .rename(columns = {\"documentation_required\": \"docu_req\"})\n",
    "    .assign(\n",
    "        num_constraints = lambda df: (\n",
    "            df[\"sdma\"] + df[\"lra\"] + (df[\"ex_parte\"] == 0) + df[\"docu_req\"]\n",
    "        ),\n",
    "        has_constraint = lambda df: df[\"num_constraints\"] > 0\n",
    "    )\n",
    "    .merge(\n",
    "        latest_totals_adj\n",
    "        [[\n",
    "            \"state\",\n",
    "            \"per_100k\",\n",
    "        ]],\n",
    "        how = \"left\",\n",
    "    )\n",
    "    .merge(\n",
    "        populations\n",
    "        [[\n",
    "            \"state\",\n",
    "            \"pop_adult\",\n",
    "        ]],\n",
    "        how = \"left\"\n",
    "    )\n",
    "    .assign(\n",
    "        pop_adult_log10 = lambda df: df[\"pop_adult\"].apply(np.log10),\n",
    "        in_sample = lambda df: df[\"state\"].isin(latest_totals[\"state\"])\n",
    "    )\n",
    ")\n",
    "\n",
    "regimes.sort_values(\"per_100k\", ascending=False).dropna()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4307d942",
   "metadata": {},
   "source": [
    "Jurisdictions outside our sample were roughly equally likely (~80%) to exhibit at least one of these four constraints (either `sdma==1`, `lra==1`, `documentation_required==1`, or `ex_parte==0`):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "b1e33cca",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:47.042873Z",
     "iopub.status.busy": "2021-09-15T18:32:47.042205Z",
     "iopub.status.idle": "2021-09-15T18:32:47.051044Z",
     "shell.execute_reply": "2021-09-15T18:32:47.051555Z"
    }
   },
   "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>has_constraint</th>\n",
       "      <th>False</th>\n",
       "      <th>True</th>\n",
       "      <th>prop</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>in_sample</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>False</th>\n",
       "      <td>6</td>\n",
       "      <td>22</td>\n",
       "      <td>0.786</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>True</th>\n",
       "      <td>4</td>\n",
       "      <td>19</td>\n",
       "      <td>0.826</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "has_constraint  False  True   prop\n",
       "in_sample                         \n",
       "False               6    22  0.786\n",
       "True                4    19  0.826"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(    \n",
    "    regimes\n",
    "    .groupby([\"in_sample\"])\n",
    "    [\"has_constraint\"]\n",
    "    .value_counts(normalize=False)\n",
    "    .unstack()\n",
    "    .assign(prop = lambda df: (df[True] / df.sum(axis = 1)).round(3))\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2a6ad029",
   "metadata": {},
   "source": [
    "... and both had a median of one constraint, though out-of-sample jurisdictions had slightly more total constraints on average:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "ea83f1ec",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:47.064291Z",
     "iopub.status.busy": "2021-09-15T18:32:47.063523Z",
     "iopub.status.idle": "2021-09-15T18:32:47.066423Z",
     "shell.execute_reply": "2021-09-15T18:32:47.066839Z"
    }
   },
   "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>c_median</th>\n",
       "      <th>c_mean</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>in_sample</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>False</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.464</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>True</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.087</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           c_median  c_mean\n",
       "in_sample                  \n",
       "False           1.0   1.464\n",
       "True            1.0   1.087"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(    \n",
    "    regimes\n",
    "    .groupby([\"in_sample\"])\n",
    "    .pipe(lambda grp: pd.DataFrame({\n",
    "        \"c_median\": grp[\"num_constraints\"].median(),\n",
    "        \"c_mean\": grp[\"num_constraints\"].mean().round(3),\n",
    "    }))\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b3a8a0fe",
   "metadata": {},
   "source": [
    "The number of constraints, however, does not seem to be clearly predictive of per-capita filing rates, especially given how few jurisdictions fall into most buckets:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "39ae4602",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:47.070083Z",
     "iopub.status.busy": "2021-09-15T18:32:47.069350Z",
     "iopub.status.idle": "2021-09-15T18:32:47.091929Z",
     "shell.execute_reply": "2021-09-15T18:32:47.092431Z"
    }
   },
   "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>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>num_constraints</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4.0</td>\n",
       "      <td>95.71</td>\n",
       "      <td>49.42</td>\n",
       "      <td>44.37</td>\n",
       "      <td>68.64</td>\n",
       "      <td>88.56</td>\n",
       "      <td>115.64</td>\n",
       "      <td>161.36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>13.0</td>\n",
       "      <td>61.36</td>\n",
       "      <td>42.53</td>\n",
       "      <td>31.49</td>\n",
       "      <td>40.43</td>\n",
       "      <td>43.09</td>\n",
       "      <td>51.60</td>\n",
       "      <td>165.62</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6.0</td>\n",
       "      <td>67.69</td>\n",
       "      <td>23.81</td>\n",
       "      <td>32.50</td>\n",
       "      <td>55.80</td>\n",
       "      <td>67.28</td>\n",
       "      <td>82.49</td>\n",
       "      <td>99.55</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 count   mean    std    min    25%    50%     75%     max\n",
       "num_constraints                                                          \n",
       "0                  4.0  95.71  49.42  44.37  68.64  88.56  115.64  161.36\n",
       "1                 13.0  61.36  42.53  31.49  40.43  43.09   51.60  165.62\n",
       "2                  6.0  67.69  23.81  32.50  55.80  67.28   82.49   99.55\n",
       "3                  0.0    NaN    NaN    NaN    NaN    NaN     NaN     NaN"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(    \n",
    "    regimes\n",
    "    .groupby(\"num_constraints\")\n",
    "    [\"per_100k\"]\n",
    "    .describe()\n",
    "    .round(2)\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cb4adaba",
   "metadata": {},
   "source": [
    "... and it is difficult to discern clear patterns from each individual constraint, in part because the comparisons each involve one very small grouping:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "bd732b1e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:47.100077Z",
     "iopub.status.busy": "2021-09-15T18:32:47.094763Z",
     "iopub.status.idle": "2021-09-15T18:32:47.134088Z",
     "shell.execute_reply": "2021-09-15T18:32:47.134606Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "=== sdma ===\n",
      "sdma        0      1\n",
      "count   21.00   2.00\n",
      "mean    68.24  76.76\n",
      "std     41.42  32.24\n",
      "min     31.49  53.96\n",
      "25%     42.26  65.36\n",
      "50%     50.22  76.76\n",
      "75%     76.73  88.16\n",
      "max    165.62  99.55\n",
      "\n",
      "=== lra ===\n",
      "lra         0       1\n",
      "count    5.00   18.00\n",
      "mean    85.19   64.48\n",
      "std     48.84   37.78\n",
      "min     43.09   31.49\n",
      "25%     44.37   40.89\n",
      "50%     76.73   50.91\n",
      "75%    100.40   70.28\n",
      "max    161.36  165.62\n",
      "\n",
      "=== ex_parte ===\n",
      "ex_parte      0       1\n",
      "count      2.00   21.00\n",
      "mean      52.19   70.58\n",
      "std       12.87   41.66\n",
      "min       43.09   31.49\n",
      "25%       47.64   42.26\n",
      "50%       52.19   51.60\n",
      "75%       56.75   85.56\n",
      "max       61.30  165.62\n",
      "\n",
      "=== docu_req ===\n",
      "docu_req       0      1\n",
      "count      20.00   3.00\n",
      "mean       69.77  63.78\n",
      "std        42.21  27.78\n",
      "min        31.49  32.50\n",
      "25%        42.77  52.88\n",
      "50%        50.91  73.27\n",
      "75%        82.44  79.42\n",
      "max       165.62  85.56\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for col in [ \"sdma\", \"lra\", \"ex_parte\", \"docu_req\" ]:\n",
    "    print(f\"=== {col} ===\")\n",
    "    print(regimes.groupby(col)[\"per_100k\"].describe().T.round(2))\n",
    "    print(\"\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "30d4b17f",
   "metadata": {},
   "source": [
    "# Appendix C: Examining spatial correlation\n",
    "\n",
    "Statisticians suggested that we also check for spatial correlation in our observations — i.e., do states near one another have a greater tendency to have similar rates?\n",
    "\n",
    "To do so, we compare each state's filing rate to those of its neighbors (states that share a border or corner):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "6f817159",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:47.141206Z",
     "iopub.status.busy": "2021-09-15T18:32:47.137190Z",
     "iopub.status.idle": "2021-09-15T18:32:47.155633Z",
     "shell.execute_reply": "2021-09-15T18:32:47.156056Z"
    }
   },
   "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>state</th>\n",
       "      <th>neighbor</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Alabama</td>\n",
       "      <td>Florida</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Alabama</td>\n",
       "      <td>Georgia</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Alabama</td>\n",
       "      <td>Mississippi</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Alabama</td>\n",
       "      <td>Tennessee</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Arkansas</td>\n",
       "      <td>Louisiana</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      state     neighbor\n",
       "0   Alabama      Florida\n",
       "1   Alabama      Georgia\n",
       "2   Alabama  Mississippi\n",
       "3   Alabama    Tennessee\n",
       "4  Arkansas    Louisiana"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "adjacencies = (\n",
    "    # Data source: https://gist.github.com/rietta/4112447\n",
    "    pd.read_csv(\"../data/raw/adjacent-states.csv\")\n",
    "    .assign(\n",
    "        state = lambda df: df[\"state\"].apply(states.lookup).astype(str),\n",
    "        neighbor = lambda df: df[\"neighbor\"].apply(states.lookup).astype(str),\n",
    "    )\n",
    ")    \n",
    "adjacencies.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "f8d0b122",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:47.162467Z",
     "iopub.status.busy": "2021-09-15T18:32:47.158493Z",
     "iopub.status.idle": "2021-09-15T18:32:47.164592Z",
     "shell.execute_reply": "2021-09-15T18:32:47.165067Z"
    }
   },
   "outputs": [],
   "source": [
    "def compare_neighbors(rates):\n",
    "    return (\n",
    "        rates\n",
    "        [[\n",
    "            \"state\",\n",
    "            \"per_100k\",\n",
    "        ]]\n",
    "        .rename(columns = { \"state\": \"neighbor\" })\n",
    "        .merge(\n",
    "            adjacencies,\n",
    "            on = \"neighbor\",\n",
    "            how = \"left\",\n",
    "        )\n",
    "        .groupby(\"state\")\n",
    "        .pipe(lambda grp: pd.DataFrame({\n",
    "            \"per_100k_neighbor_mean\": grp[\"per_100k\"].mean(),\n",
    "            \"neighbors\": grp[\"neighbor\"].apply(set).apply(sorted),\n",
    "        }))\n",
    "        .reset_index()\n",
    "        .pipe(lambda df: (\n",
    "            rates\n",
    "            [[\n",
    "                \"state\",\n",
    "                \"per_100k\",\n",
    "            ]]\n",
    "            .merge(\n",
    "                df,\n",
    "                how = \"left\",\n",
    "                on = \"state\",\n",
    "            )\n",
    "        ))\n",
    "        .dropna()\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "64f97c5a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:47.192483Z",
     "iopub.status.busy": "2021-09-15T18:32:47.167427Z",
     "iopub.status.idle": "2021-09-15T18:32:47.195600Z",
     "shell.execute_reply": "2021-09-15T18:32:47.196030Z"
    }
   },
   "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>state</th>\n",
       "      <th>per_100k</th>\n",
       "      <th>per_100k_neighbor_mean</th>\n",
       "      <th>neighbors</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Arizona</td>\n",
       "      <td>42.263156</td>\n",
       "      <td>36.870162</td>\n",
       "      <td>[Colorado, New Mexico, Utah]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Arkansas</td>\n",
       "      <td>50.223308</td>\n",
       "      <td>61.298432</td>\n",
       "      <td>[Missouri]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Colorado</td>\n",
       "      <td>31.494369</td>\n",
       "      <td>63.519803</td>\n",
       "      <td>[Arizona, Kansas, Nebraska, New Mexico, Utah]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Connecticut</td>\n",
       "      <td>165.618513</td>\n",
       "      <td>73.271076</td>\n",
       "      <td>[Massachusetts]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>District of Columbia</td>\n",
       "      <td>99.553136</td>\n",
       "      <td>37.724706</td>\n",
       "      <td>[Virginia]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  state    per_100k  per_100k_neighbor_mean  \\\n",
       "0               Arizona   42.263156               36.870162   \n",
       "1              Arkansas   50.223308               61.298432   \n",
       "2              Colorado   31.494369               63.519803   \n",
       "3           Connecticut  165.618513               73.271076   \n",
       "4  District of Columbia   99.553136               37.724706   \n",
       "\n",
       "                                       neighbors  \n",
       "0                   [Colorado, New Mexico, Utah]  \n",
       "1                                     [Missouri]  \n",
       "2  [Arizona, Kansas, Nebraska, New Mexico, Utah]  \n",
       "3                                [Massachusetts]  \n",
       "4                                     [Virginia]  "
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "neighbor_comparison = compare_neighbors(latest_totals)\n",
    "\n",
    "neighbor_comparison.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c2e1595a",
   "metadata": {},
   "source": [
    "We see some correlation, but it does not reach the standard level of statistical significance — although, again, our sample size may simply be too small to detect it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "9a73e837",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:47.227774Z",
     "iopub.status.busy": "2021-09-15T18:32:47.221072Z",
     "iopub.status.idle": "2021-09-15T18:32:47.459574Z",
     "shell.execute_reply": "2021-09-15T18:32:47.460187Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "(\n",
    "    neighbor_comparison\n",
    "    .pipe(lambda df: sns.regplot(\n",
    "        y = df[\"per_100k\"],\n",
    "        x = df[\"per_100k_neighbor_mean\"],\n",
    "        truncate = False,\n",
    "        scatter_kws = { \"alpha\": 0.5 },\n",
    "    ))\n",
    ")\n",
    "pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "e1e6171e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2021-09-15T18:32:47.477792Z",
     "iopub.status.busy": "2021-09-15T18:32:47.477083Z",
     "iopub.status.idle": "2021-09-15T18:32:47.480059Z",
     "shell.execute_reply": "2021-09-15T18:32:47.480511Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<caption>OLS Regression Results</caption>\n",
       "<tr>\n",
       "  <th>Dep. Variable:</th>        <td>per_100k</td>     <th>  R-squared:         </th> <td>   0.056</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model:</th>                   <td>OLS</td>       <th>  Adj. R-squared:    </th> <td>   0.011</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Method:</th>             <td>Least Squares</td>  <th>  F-statistic:       </th> <td>   1.253</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Date:</th>             <td>Fri, 17 Sep 2021</td> <th>  Prob (F-statistic):</th>  <td> 0.276</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Time:</th>                 <td>10:23:26</td>     <th>  Log-Likelihood:    </th> <td> -116.40</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>No. Observations:</th>      <td>    23</td>      <th>  AIC:               </th> <td>   236.8</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Df Residuals:</th>          <td>    21</td>      <th>  BIC:               </th> <td>   239.1</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Df Model:</th>              <td>     1</td>      <th>                     </th>     <td> </td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Covariance Type:</th>      <td>nonrobust</td>    <th>                     </th>     <td> </td>   \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "             <td></td>               <th>coef</th>     <th>std err</th>      <th>t</th>      <th>P>|t|</th>  <th>[0.025</th>    <th>0.975]</th>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>const</th>                  <td>   39.2511</td> <td>   27.833</td> <td>    1.410</td> <td> 0.173</td> <td>  -18.631</td> <td>   97.133</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>per_100k_neighbor_mean</th> <td>    0.4742</td> <td>    0.424</td> <td>    1.120</td> <td> 0.276</td> <td>   -0.407</td> <td>    1.355</td>\n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "  <th>Omnibus:</th>       <td> 9.862</td> <th>  Durbin-Watson:     </th> <td>   1.961</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(Omnibus):</th> <td> 0.007</td> <th>  Jarque-Bera (JB):  </th> <td>   8.080</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Skew:</th>          <td> 1.402</td> <th>  Prob(JB):          </th> <td>  0.0176</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Kurtosis:</th>      <td> 3.757</td> <th>  Cond. No.          </th> <td>    220.</td>\n",
       "</tr>\n",
       "</table><br/><br/>Notes:<br/>[1] Standard Errors assume that the covariance matrix of the errors is correctly specified."
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary.Summary'>\n",
       "\"\"\"\n",
       "                            OLS Regression Results                            \n",
       "==============================================================================\n",
       "Dep. Variable:               per_100k   R-squared:                       0.056\n",
       "Model:                            OLS   Adj. R-squared:                  0.011\n",
       "Method:                 Least Squares   F-statistic:                     1.253\n",
       "Date:                Fri, 17 Sep 2021   Prob (F-statistic):              0.276\n",
       "Time:                        10:23:26   Log-Likelihood:                -116.40\n",
       "No. Observations:                  23   AIC:                             236.8\n",
       "Df Residuals:                      21   BIC:                             239.1\n",
       "Df Model:                           1                                         \n",
       "Covariance Type:            nonrobust                                         \n",
       "==========================================================================================\n",
       "                             coef    std err          t      P>|t|      [0.025      0.975]\n",
       "------------------------------------------------------------------------------------------\n",
       "const                     39.2511     27.833      1.410      0.173     -18.631      97.133\n",
       "per_100k_neighbor_mean     0.4742      0.424      1.120      0.276      -0.407       1.355\n",
       "==============================================================================\n",
       "Omnibus:                        9.862   Durbin-Watson:                   1.961\n",
       "Prob(Omnibus):                  0.007   Jarque-Bera (JB):                8.080\n",
       "Skew:                           1.402   Prob(JB):                       0.0176\n",
       "Kurtosis:                       3.757   Cond. No.                         220.\n",
       "==============================================================================\n",
       "\n",
       "Notes:\n",
       "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
       "\"\"\""
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(\n",
    "    neighbor_comparison\n",
    "    .pipe(lambda df: (\n",
    "        sm.OLS(\n",
    "            df[\"per_100k\"],\n",
    "            sm.add_constant(df[\"per_100k_neighbor_mean\"]),\n",
    "        )\n",
    "     ))\n",
    "    .fit()\n",
    "    .summary()\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "45e3fd99",
   "metadata": {},
   "source": [
    "---\n",
    "\n",
    "---\n",
    "\n",
    "---"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.9.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}