taxcalc 4.2.1__py3-none-any.whl → 4.3.0__py3-none-any.whl

taxcalc/tests/test_policy.py

@@ -0,0 +1,1557 @@
+"""
+Test Policy class and its methods.
+"""
+# CODING-STYLE CHECKS:
+# pycodestyle test_policy.py
+# pylint --disable=locally-disabled test_policy.py
+#
+# pylint: disable=too-many-lines
+
+import copy
+import os
+import json
+import numpy as np
+import pytest
+import paramtools as pt
+# pylint: disable=import-error
+from taxcalc import Policy
+
+
+def cmp_policy_objs(pol1, pol2, year_range=None, exclude=None):
+    """
+    Compare parameter values two policy objects.
+
+    year_range: years over which to compare values.
+    exclude: list of parameters to exclude from comparison.
+    """
+    if year_range is not None:
+        pol1.set_state(year=list(year_range))
+        pol2.set_state(year=list(year_range))
+    else:
+        pol1.clear_state()
+        pol2.clear_state()
+    for param in pol1._data:
+        if exclude and param in exclude:
+            continue
+        v1 = getattr(pol1, param)
+        v2 = getattr(pol2, param)
+        np.testing.assert_allclose(v1, v2)
+
+
+def test_incorrect_class_instantiation():
+    """
+    Test incorrect instantiation of Policy class object.
+    """
+    with pytest.raises(ValueError):
+        Policy(gfactors=[])
+
+
+def test_correct_class_instantiation():
+    """
+    Test correct instantiation of Policy class object.
+    """
+    pol = Policy()
+    assert pol
+    pol.implement_reform({})
+    with pytest.raises(pt.ValidationError):
+        pol.implement_reform([])
+    with pytest.raises(pt.ValidationError):
+        pol.implement_reform({2099: {'II_em': 99000}})
+    pol.set_year(2019)
+    with pytest.raises(pt.ValidationError):
+        pol.implement_reform({2018: {'II_em': 99000}})
+    with pytest.raises(pt.ValidationError):
+        pol.implement_reform({2020: {'II_em': -1000}})
+
+
+def test_json_reform_url():
+    """
+    Test reading a JSON reform from a URL. Results from the URL are expected
+    to match the results from the string.
+    """
+    reform_str = """
+    {
+        // raise FICA payroll tax rate in 2018 and 2020
+        "FICA_ss_trt_employer": {
+            "2018": 0.065,
+            "2020": 0.070
+        },
+        "FICA_ss_trt_employee": {
+            "2018": 0.065,
+            "2020": 0.070
+        },
+        // raise Medicare payroll tax rate in 2019 and 2021
+        "FICA_mc_trt_employer": {
+            "2019": 0.015,
+            "2021": 0.016
+        },
+        "FICA_mc_trt_employee": {
+            "2019": 0.015,
+            "2021": 0.016
+        }
+    }
+    """
+    reform_url = ('https://raw.githubusercontent.com/PSLmodels/'
+                  'Tax-Calculator/master/taxcalc/reforms/ptaxes0.json')
+    params_str = Policy.read_json_reform(reform_str)
+    params_url = Policy.read_json_reform(reform_url)
+    assert params_str == params_url
+
+    reform_gh_url = (
+        "github://PSLmodels:Tax-Calculator@master/taxcalc/reforms/ptaxes0.json"
+    )
+    params_gh_url = Policy.read_json_reform(reform_gh_url)
+    assert params_gh_url
+    assert params_gh_url == params_str
+
+
+REFORM_JSON = """
+// Example of a reform file suitable for Policy.read_json_reform().
+// This JSON file can contain any number of trailing //-style comments, which
+// will be removed before the contents are converted from JSON to a dictionary.
+// The primary keys are parameters and the secondary keys are years.
+// Both the primary and secondary key values must be enclosed in quotes (").
+// Boolean variables are specified as true or false (no quotes; all lowercase).
+{
+    "AMT_brk1": // top of first AMT tax bracket
+    {"2015": 200000,
+     "2017": 300000
+    },
+    "EITC_c": // maximum EITC amount by number of qualifying kids (0,1,2,3+)
+    {"2016": [ 900, 5000,  8000,  9000],
+     "2019": [1200, 7000, 10000, 12000]
+    },
+    "II_em": // personal exemption amount (see indexing changes below)
+    {"2016": 6000,
+     "2018": 7500,
+     "2020": 9000
+    },
+    "II_em-indexed": // personal exemption amount indexing status
+    {"2016": false, // values in future years are same as this year value
+     "2018": true   // values in future years indexed with this year as base
+    },
+    "SS_Earnings_c": // social security (OASDI) maximum taxable earnings
+    {"2016": 300000,
+     "2018": 500000,
+     "2020": 700000
+    },
+    "AMT_em-indexed": // AMT exemption amount indexing status
+    {"2017": false, // values in future years are same as this year value
+     "2020": true   // values in future years indexed with this year as base
+    }
+}
+"""
+
+
+# pylint: disable=protected-access,no-member
+
+
+@pytest.mark.parametrize("set_year", [False, True])
+def test_read_json_reform_file_and_implement_reform(set_year):
+    """
+    Test reading and translation of reform JSON into a reform dictionary
+    and then using that reform dictionary to implement reform.
+    """
+    pol = Policy()
+    if set_year:
+        pol.set_year(2015)
+    pol.implement_reform(Policy.read_json_reform(REFORM_JSON))
+    syr = pol.start_year
+    # pylint: disable=protected-access
+    amt_brk1 = pol._AMT_brk1
+    assert amt_brk1[2015 - syr] == 200000
+    assert amt_brk1[2016 - syr] > 200000
+    assert amt_brk1[2017 - syr] == 300000
+    assert amt_brk1[2018 - syr] > 300000
+    ii_em = pol._II_em
+    assert ii_em[2016 - syr] == 6000
+    assert ii_em[2017 - syr] == 6000
+    assert ii_em[2018 - syr] == 7500
+    assert ii_em[2019 - syr] > 7500
+    assert ii_em[2020 - syr] == 9000
+    assert ii_em[2021 - syr] > 9000
+    amt_em = pol._AMT_em
+    assert amt_em[2016 - syr, 0] > amt_em[2015 - syr, 0]
+    assert amt_em[2017 - syr, 0] > amt_em[2016 - syr, 0]
+    assert amt_em[2018 - syr, 0] == amt_em[2017 - syr, 0]
+    assert amt_em[2019 - syr, 0] == amt_em[2017 - syr, 0]
+    assert amt_em[2020 - syr, 0] == amt_em[2017 - syr, 0]
+    assert amt_em[2021 - syr, 0] > amt_em[2020 - syr, 0]
+    assert amt_em[2022 - syr, 0] > amt_em[2021 - syr, 0]
+    add4aged = pol._ID_Medical_frt_add4aged
+    assert add4aged[2015 - syr] == -0.025
+    assert add4aged[2016 - syr] == -0.025
+    assert add4aged[2017 - syr] == 0.0
+    assert add4aged[2022 - syr] == 0.0
+
+
+def test_constant_inflation_rate_with_reform():
+    """
+    Test indexing of policy parameters involved in a reform.
+    """
+    pol = Policy()
+    # implement reform in year before final year
+    fyr = 2034
+    ryr = fyr - 1
+    reform = {
+        'II_em': {(ryr - 3): 1000,  # to avoid divide-by-zero under TCJA
+                  ryr: 20000}
+    }
+    pol.implement_reform(reform)
+    # extract price inflation rates
+    pirates = pol.inflation_rates()
+    syr = Policy.JSON_START_YEAR
+    irate_b = pirates[ryr - 2 - syr]
+    irate_a = pirates[ryr - syr]
+    # check implied inflation rate just before reform
+    grate = float(pol._II_em[ryr - 1 - syr]) / float(pol._II_em[ryr - 2 - syr])
+    assert round(grate - 1.0, 4) == round(irate_b, 4)
+    # check implied inflation rate just after reform
+    grate = float(pol._II_em[ryr + 1 - syr]) / float(pol._II_em[ryr - syr])
+    assert round(grate - 1.0, 6) == round(irate_a, 6)
+
+
+def test_variable_inflation_rate_with_reform():
+    """
+    Test indexing of policy parameters involved in a reform.
+    """
+    pol = Policy()
+    syr = Policy.JSON_START_YEAR
+    assert pol._II_em[2013 - syr] == 3900
+    # implement reform in 2020 which is two years before the last year, 2022
+    reform = {
+        'II_em': {2018: 1000,  # to avoid divide-by-zero under TCJA
+                  2020: 20000}
+    }
+    pol.implement_reform(reform)
+    pol.set_year(2020)
+    assert pol.current_year == 2020
+    # extract price inflation rates
+    pirates = pol.inflation_rates()
+    irate2018 = pirates[2018 - syr]
+    irate2020 = pirates[2020 - syr]
+    irate2021 = pirates[2021 - syr]
+    # check implied inflation rate between 2018 and 2019 (before the reform)
+    grate = float(pol._II_em[2019 - syr]) / float(pol._II_em[2018 - syr])
+    assert round(grate - 1.0, 5) == round(irate2018, 5)
+    # check implied inflation rate between 2020 and 2021 (after the reform)
+    grate = float(pol._II_em[2021 - syr]) / float(pol._II_em[2020 - syr])
+    assert round(grate - 1.0, 5) == round(irate2020, 5)
+    # check implied inflation rate between 2021 and 2022 (after the reform)
+    grate = float(pol._II_em[2022 - syr]) / float(pol._II_em[2021 - syr])
+    assert round(grate - 1.0, 5) == round(irate2021, 5)
+
+
+def test_multi_year_reform():
+    """
+    Test multi-year reform involving 1D and 2D parameters.
+    """
+    # specify dimensions of policy Policy object
+    syr = Policy.JSON_START_YEAR
+    nyrs = Policy.DEFAULT_NUM_YEARS
+    pol = Policy()
+    iratelist = pol.inflation_rates()
+    ifactor = {}
+    for i in range(0, nyrs):
+        ifactor[syr + i] = 1.0 + iratelist[i]
+    wratelist = pol.wage_growth_rates()
+    wfactor = {}
+    for i in range(0, nyrs):
+        wfactor[syr + i] = 1.0 + wratelist[i]
+    # specify multi-year reform using a param:year:value-fomatted dictionary
+    reform = {
+        'SS_Earnings_c': {2016: 300000,
+                          2017: 500000,
+                          2019: 700000},
+        'SS_Earnings_c-indexed': {2017: False,
+                                  2019: True},
+        'CTC_c': {2015: 2000},
+        'EITC_c': {2016: [900, 5000, 8000, 9000],
+                   2019: [1200, 7000, 10000, 12000]},
+        'II_em': {2016: 7000,
+                  2019: 9000}
+    }
+    # implement multi-year reform
+    pol.implement_reform(reform)
+    assert pol.current_year == syr
+    # move policy Policy object forward in time so current_year is syr+2
+    #   Note: this would be typical usage because the first budget year
+    #         is typically greater than Policy start_year.
+    pol.set_year(pol.start_year + 2)
+    assert pol.current_year == syr + 2
+    # confirm that actual parameters have expected post-reform values
+    check_eitc_c(pol, reform, ifactor)
+    check_ii_em(pol, reform, ifactor)
+    check_ss_earnings_c(pol, reform, wfactor)
+    check_ctc_c(pol, reform)
+    # end of test_multi_year_reform with the check_* functions below:
+
+
+def check_ctc_c(ppo, reform):
+    """
+    Compare actual and expected _CTC_c parameter values
+    generated by the test_multi_year_reform() function above.
+    Ensure that future-year values in policy_current_law.json
+    are overwritten by reform.
+    """
+    actual = {}
+    arr = getattr(ppo, '_CTC_c')
+    for i in range(0, ppo.num_years):
+        actual[ppo.start_year + i] = arr[i]
+    assert actual[2013] == 1000
+    assert actual[2014] == 1000
+    e2015 = reform['CTC_c'][2015]
+    assert actual[2015] == e2015
+    e2016 = actual[2015]
+    assert actual[2016] == e2016
+    e2017 = actual[2016]
+    assert actual[2017] == e2017
+    e2018 = actual[2017]
+    assert actual[2018] == e2018
+    e2019 = actual[2018]
+    assert actual[2019] == e2019
+
+
+def check_eitc_c(ppo, reform, ifactor):
+    """
+    Compare actual and expected _EITC_c parameter values
+    generated by the test_multi_year_reform() function above.
+    """
+    actual = {}
+    arr = getattr(ppo, '_EITC_c')
+    alen = len(arr[0])
+    for i in range(0, ppo.num_years):
+        actual[ppo.start_year + i] = arr[i]
+    assert np.allclose(actual[2013], [487, 3250, 5372, 6044],
+                       atol=0.01, rtol=0.0)
+    assert np.allclose(actual[2014], [496, 3305, 5460, 6143],
+                       atol=0.01, rtol=0.0)
+    assert np.allclose(actual[2015], [503, 3359, 5548, 6242],
+                       atol=0.01, rtol=0.0)
+    e2016 = reform['EITC_c'][2016]
+    assert np.allclose(actual[2016], e2016, atol=0.01, rtol=0.0)
+    e2017 = [ifactor[2016] * actual[2016][j] for j in range(0, alen)]
+    assert np.allclose(actual[2017], e2017, atol=0.01, rtol=0.0)
+    e2018 = [ifactor[2017] * actual[2017][j] for j in range(0, alen)]
+    assert np.allclose(actual[2018], e2018, atol=0.01, rtol=0.0)
+    e2019 = reform['EITC_c'][2019]
+    assert np.allclose(actual[2019], e2019, atol=0.01, rtol=0.0)
+    e2020 = [ifactor[2019] * actual[2019][j] for j in range(0, alen)]
+    assert np.allclose(actual[2020], e2020, atol=0.01, rtol=0.0)
+    e2021 = [ifactor[2020] * actual[2020][j] for j in range(0, alen)]
+    assert np.allclose(actual[2021], e2021, atol=0.01, rtol=0.0)
+    e2022 = [ifactor[2021] * actual[2021][j] for j in range(0, alen)]
+    assert np.allclose(actual[2022], e2022, atol=0.01, rtol=0.0)
+
+
+def check_ii_em(ppo, reform, ifactor):
+    """
+    Compare actual and expected _II_em parameter values
+    generated by the test_multi_year_reform() function above.
+    """
+    actual = {}
+    arr = getattr(ppo, '_II_em')
+    for i in range(0, ppo.num_years):
+        actual[ppo.start_year + i] = arr[i]
+    assert actual[2013] == 3900
+    assert actual[2014] == 3950
+    assert actual[2015] == 4000
+    e2016 = reform['II_em'][2016]
+    assert actual[2016] == e2016
+    e2017 = ifactor[2016] * actual[2016]
+    assert np.allclose([actual[2017]], [e2017], atol=0.01, rtol=0.0)
+    e2018 = ifactor[2017] * actual[2017]
+    assert np.allclose([actual[2018]], [e2018], atol=0.01, rtol=0.0)
+    e2019 = reform['II_em'][2019]
+    assert actual[2019] == e2019
+    e2020 = ifactor[2019] * actual[2019]
+    assert np.allclose([actual[2020]], [e2020], atol=0.01, rtol=0.0)
+    e2021 = ifactor[2020] * actual[2020]
+    assert np.allclose([actual[2021]], [e2021], atol=0.01, rtol=0.0)
+    e2022 = ifactor[2021] * actual[2021]
+    assert np.allclose([actual[2022]], [e2022], atol=0.01, rtol=0.0)
+
+
+def check_ss_earnings_c(ppo, reform, wfactor):
+    """
+    Compare actual and expected _SS_Earnings_c parameter values
+    generated by the test_multi_year_reform() function above.
+    """
+    actual = {}
+    arr = getattr(ppo, '_SS_Earnings_c')
+    for i in range(0, ppo.num_years):
+        actual[ppo.start_year + i] = arr[i]
+    assert actual[2013] == 113700
+    assert actual[2014] == 117000
+    assert actual[2015] == 118500
+    e2016 = reform['SS_Earnings_c'][2016]
+    assert actual[2016] == e2016
+    e2017 = reform['SS_Earnings_c'][2017]
+    assert actual[2017] == e2017
+    e2018 = actual[2017]  # no indexing after 2017
+    assert actual[2018] == e2018
+    e2019 = reform['SS_Earnings_c'][2019]
+    assert actual[2019] == e2019
+    e2020 = wfactor[2019] * actual[2019]  # indexing after 2019
+    assert np.allclose([actual[2020]], [e2020], atol=0.01, rtol=0.0)
+    e2021 = wfactor[2020] * actual[2020]
+    assert np.allclose([actual[2021]], [e2021], atol=0.01, rtol=0.0)
+    e2022 = wfactor[2021] * actual[2021]
+    assert np.allclose([actual[2022]], [e2022], atol=0.01, rtol=0.0)
+
+
+def test_policy_metadata():
+    """
+    Test that metadata() method returns expected dictionary.
+    """
+    clp = Policy()
+    mdata = clp.metadata()
+    assert mdata
+
+
+def test_implement_reform_raises_on_no_year():
+    """
+    Test that implement_reform raises error for missing year.
+    """
+    reform = {'STD_Aged': [1400, 1200, 1400, 1400, 1400]}
+    ppo = Policy()
+    with pytest.raises(pt.ValidationError):
+        ppo.implement_reform(reform)
+
+
+def test_implement_reform_raises_on_early_year():
+    """
+    Test that implement_reform raises error for early year.
+    """
+    ppo = Policy()
+    reform = {'STD_Aged': {2010: [1400, 1100, 1100, 1400, 1400]}}
+    with pytest.raises(pt.ValidationError):
+        ppo.implement_reform(reform)
+
+
+def test_reform_with_default_indexed():
+    """
+    Test that implement_reform indexes after first reform year.
+    """
+    ppo = Policy()
+    reform = {'II_em': {2015: 4300}}
+    ppo.implement_reform(reform)
+    # II_em has a default indexing status of true, so
+    # in 2016 its value should be greater than 4300
+    ppo.set_year(2016)
+    assert ppo.II_em > 4300
+
+
+def test_reform_makes_no_changes_before_year():
+    """
+    Test that implement_reform makes no changes before first reform year.
+    """
+    ppo = Policy()
+    reform = {'II_em': {2015: 4400}, 'II_em-indexed': {2015: True}}
+    ppo.implement_reform(reform)
+    ppo.set_year(2015)
+    assert np.allclose(ppo._II_em[:3], np.array([3900, 3950, 4400]),
+                       atol=0.01, rtol=0.0)
+    assert ppo.II_em == 4400
+
+
+@pytest.mark.parametrize("set_year", [False, True])
+def test_read_json_reform_and_implement_reform(set_year):
+    """
+    Test reading and translation of reform file into a reform dictionary
+    that is then used to call implement_reform method.
+    NOTE: implement_reform called when policy.current_year == policy.start_year
+    """
+    reform_json = """
+    // Example of JSON reform text suitable for the
+    // Policy.read_json_reform() method.
+    // This JSON text can contain any number of trailing //-style comments,
+    // which will be removed before the contents are converted from JSON to
+    // a dictionary.
+    // The primary keys are policy parameters and secondary keys are years.
+    // Both the primary & secondary key values must be enclosed in quotes (").
+    // Boolean variables are specified as true or false with no quotes and all
+    // lowercase characters.
+    {
+        "AMT_brk1": // top of first AMT tax bracket
+        {"2015": 200000,
+         "2017": 300000
+        },
+        "EITC_c": // max EITC amount by number of qualifying kids (0,1,2,3+)
+        {"2016": [ 900, 5000,  8000,  9000],
+         "2019": [1200, 7000, 10000, 12000]
+        },
+        "II_em": // personal exemption amount (see indexing changes below)
+        {"2016": 6000,
+         "2018": 7500,
+         "2020": 9000
+        },
+        "II_em-indexed": // personal exemption amount indexing status
+        {"2016": false, // values in future years are same as this year value
+         "2018": true   // vals in future years indexed with this year as base
+        },
+        "SS_Earnings_c": // Social Security (OASDI) maximum taxable earnings
+        {"2016": 300000,
+         "2018": 500000,
+         "2020": 700000
+        },
+        "AMT_em-indexed": // AMT exemption amount indexing status
+        {"2017": false, // values in future years are same as this year value
+         "2020": true   // vals in future years indexed with this year as base
+        }
+    }
+    """
+    policy = Policy()
+    if set_year:
+        policy.set_year(2015)
+    reform_dict = Policy.read_json_reform(reform_json)
+    policy.implement_reform(reform_dict)
+    syr = policy.start_year
+    amt_brk1 = policy._AMT_brk1
+    assert amt_brk1[2015 - syr] == 200000
+    assert amt_brk1[2016 - syr] > 200000
+    assert amt_brk1[2017 - syr] == 300000
+    assert amt_brk1[2018 - syr] > 300000
+    ii_em = policy._II_em
+    assert ii_em[2016 - syr] == 6000
+    assert ii_em[2017 - syr] == 6000
+    assert ii_em[2018 - syr] == 7500
+    assert ii_em[2019 - syr] > 7500
+    assert ii_em[2020 - syr] == 9000
+    assert ii_em[2021 - syr] > 9000
+    amt_em = policy._AMT_em
+    assert amt_em[2016 - syr, 0] > amt_em[2015 - syr, 0]
+    assert amt_em[2017 - syr, 0] > amt_em[2016 - syr, 0]
+    assert amt_em[2018 - syr, 0] == amt_em[2017 - syr, 0]
+    assert amt_em[2019 - syr, 0] == amt_em[2017 - syr, 0]
+    assert amt_em[2020 - syr, 0] == amt_em[2017 - syr, 0]
+    assert amt_em[2021 - syr, 0] > amt_em[2020 - syr, 0]
+    assert amt_em[2022 - syr, 0] > amt_em[2021 - syr, 0]
+    add4aged = policy._ID_Medical_frt_add4aged
+    assert add4aged[2015 - syr] == -0.025
+    assert add4aged[2016 - syr] == -0.025
+    assert add4aged[2017 - syr] == 0.0
+    assert add4aged[2022 - syr] == 0.0
+
+
+def test_pop_the_cap_reform():
+    """
+    Test eliminating the maximum taxable earnings (MTE)
+    used in the calculation of the OASDI payroll tax.
+    """
+    # create Policy parameters object
+    ppo = Policy()
+    assert ppo.current_year == Policy.JSON_START_YEAR
+    # confirm that MTE has current-law values in 2015 and 2016
+    mte = ppo._SS_Earnings_c
+    syr = Policy.JSON_START_YEAR
+    assert mte[2015 - syr] == 118500
+    assert mte[2016 - syr] == 118500
+    # specify a "pop the cap" reform that eliminates MTE cap in 2016
+    reform = {'SS_Earnings_c': {2016: 9e99}}
+    ppo.implement_reform(reform)
+    mte = ppo._SS_Earnings_c
+    assert mte[2015 - syr] == 118500
+    assert mte[2016 - syr] == 9e99
+    assert mte[ppo.end_year - syr] == 9e99
+
+
+def test_order_of_indexing_and_level_reforms():
+    """
+    Test that the order of the two reform provisions for the same parameter
+    make no difference to the post-reform policy parameter values.
+    """
+    # specify two reforms that raises the MTE and stops its indexing in 2015
+    reform = [
+        {
+            'SS_Earnings_c': {2015: 500000},
+            'SS_Earnings_c-indexed': {2015: False}
+        },
+        # now reverse the order of the two reform provisions
+        {
+            'SS_Earnings_c-indexed': {2015: False},
+            'SS_Earnings_c': {2015: 500000}
+        }
+    ]
+    # specify two Policy objects
+    ppo = [Policy(), Policy()]
+    # apply reforms to corresponding Policy object & check post-reform values
+    syr = Policy.JSON_START_YEAR
+    for ref in range(len(reform)):  # pylint: disable=consider-using-enumerate
+        # confirm pre-reform MTE values in 2014-2017
+        mte = ppo[ref]._SS_Earnings_c
+        assert mte[2014 - syr] == 117000
+        assert mte[2015 - syr] == 118500
+        assert mte[2016 - syr] == 118500
+        assert mte[2017 - syr] < 500000
+        # implement reform in 2015
+        ppo[ref].implement_reform(reform[ref])
+        # confirm post-reform MTE values in 2014-2017
+        mte = ppo[ref]._SS_Earnings_c
+        assert mte[2014 - syr] == 117000
+        assert mte[2015 - syr] == 500000
+        assert mte[2016 - syr] == 500000
+        assert mte[2017 - syr] == 500000
+
+
+def test_misspecified_reform_dictionary():
+    """
+    Demonstrate pitfalls of careless specification of policy reform
+    dictionaries involving non-unique dictionary keys.
+    """
+    # specify apparently the same reform in two different ways, forgetting
+    # that Python dictionaries have unique keys
+    reform1 = {'II_em': {2019: 1000, 2020: 2000}}
+    # pylint: disable=duplicate-key
+    reform2 = {'II_em': {2019: 1000}, 'II_em': {2020: 2000}}
+    # these two reform dictionaries are not the same: the second
+    # 'II_em' key value for 2020 in reform2 OVERWRITES and REPLACES
+    # the first 'II_em' key value for 2019 in reform2
+    assert reform1 != reform2
+
+
+def test_section_titles(tests_path):
+    """
+    Check section titles in policy_current_law.json and uguide.htmx files.
+    """
+    # pylint: disable=too-many-locals
+    def generate_section_dictionary(md_text):
+        """
+        Returns dictionary of section titles that is
+        structured like the VALID_SECTION dictionary (see below) and
+        extracted from the specified html_text.
+        """
+        sdict = {}
+        for line in md_text.splitlines():
+            # This is shown as an empty case in current law policy and
+            # validation.
+            if line.startswith('## Other Parameters (not in Tax-Brain webapp'):
+                sdict[''] = {}
+                sdict[''][''] = 0
+                continue
+            sec2line = line.startswith('### ')
+            sec1line = line.startswith('## ')
+            # Create outer-layer dictionary entry for sec1.
+            if sec1line:
+                sec1 = line.replace('##', '', 1).strip()
+                sdict[sec1] = {}
+            # Create inner dictionary entry for sec1-sec2.
+            # Note that sec1 will have been defined from a previous loop.
+            if sec2line:
+                sec2 = line.replace('###', '', 1).strip()
+                sdict[sec1][sec2] = 0
+        return sdict
+    # begin main logic of test_section_titles
+    # specify expected section titles ordered as on the Tax-Brain webapp
+    ided_ceiling_pct = ('Ceiling On The Benefit Of Itemized Deductions '
+                        'As A Percent Of Deductible Expenses')
+    cgqd_tax_same = ('Tax All Capital Gains And Dividends The Same '
+                     'As Regular Taxable Income')
+    valid_dict = {
+        '': {  # empty section_1 implies parameter not displayed in Tax-Brain
+            '': 0
+        },
+        'Parameter Indexing': {
+            'Offsets': 0
+        },
+        'Payroll Taxes': {
+            'Social Security FICA': 0,
+            'Medicare FICA': 0,
+            'Additional Medicare FICA': 0
+        },
+        'Social Security Taxability': {
+            'Social Security Benefit Taxability': 0,
+        },
+        'Above The Line Deductions': {
+            'Misc. Adjustment Haircuts': 0,
+            'Misc. Exclusions': 0,
+            'Child And Elderly Care': 0
+        },
+        'Personal Exemptions': {
+            'Personal And Dependent Exemption Amount': 0,
+            # 'Personal Exemption Phaseout Starting Income': 0,
+            'Personal Exemption Phaseout Rate': 0,
+            'Repeal for Dependents Under Age 18': 0
+        },
+        'Standard Deduction': {
+            'Standard Deduction Amount': 0,
+            'Additional Standard Deduction For Blind And Aged': 0
+            # 'Standard Deduction For Dependents': 0
+        },
+        'Nonrefundable Credits': {
+            'Misc. Credit Limits': 0,
+            'Child And Dependent Care': 0,
+            'Personal Nonrefundable Credit': 0
+        },
+        'Child/Dependent Credits': {
+            'Child Tax Credit': 0,
+            'Additional Child Tax Credit': 0,
+            'Other Dependent Tax Credit': 0
+        },
+        'Itemized Deductions': {
+            'Medical Expenses': 0,
+            'State And Local Income And Sales Taxes': 0,
+            'State, Local, And Foreign Real Estate Taxes': 0,
+            'State And Local Taxes And Real Estate Taxes': 0,
+            'Interest Paid': 0,
+            'Charity': 0,
+            'Casualty': 0,
+            'Miscellaneous': 0,
+            'Itemized Deduction Limitation': 0,
+            'Surtax On Itemized Deduction Benefits Above An AGI Threshold': 0,
+            ided_ceiling_pct: 0,
+            'Ceiling On The Amount Of Itemized Deductions Allowed': 0
+        },
+        'Capital Gains And Dividends': {
+            'Regular - Long Term Capital Gains And Qualified Dividends': 0,
+            'AMT - Long Term Capital Gains And Qualified Dividends': 0,
+            cgqd_tax_same: 0
+        },
+        'Personal Income': {
+            'Regular: Non-AMT, Non-Pass-Through': 0,
+            'Pass-Through': 0,
+            'Alternative Minimum Tax': 0
+        },
+        'Other Taxes': {
+            'Net Investment Income Tax': 0
+        },
+        'Refundable Credits': {
+            'Earned Income Tax Credit': 0,
+            'New Refundable Child Tax Credit': 0,
+            'Personal Refundable Credit': 0,
+            'Refundable Payroll Tax Credit': 0
+        },
+        'Surtaxes': {
+            'New Minimum Tax': 0,
+            'New AGI Surtax': 0,
+            'Lump-Sum Tax': 0
+        },
+        'Universal Basic Income': {
+            'UBI Benefits': 0,
+            'UBI Taxability': 0
+        },
+        'Benefits': {
+            'Benefit Repeal': 0,
+        }
+    }
+    # check validity of parameter section titles in policy_current_law.json
+    path = os.path.join(tests_path, '..', 'policy_current_law.json')
+    with open(path, 'r', encoding='utf-8') as clpfile:
+        clpdict = json.load(clpfile)
+        clpdict.pop("schema", None)
+    # ... make sure ever clpdict section title is in valid_dict
+    clp_dict = {}  # dictionary of clp section titles structured like valid
+    for pname in clpdict:
+        param = clpdict[pname]
+        assert isinstance(param, dict)
+        sec1title = param['section_1']
+        assert sec1title in valid_dict
+        sec2title = param['section_2']
+        assert sec2title in valid_dict[sec1title]
+        if sec1title not in clp_dict:
+            clp_dict[sec1title] = {}
+        if sec2title not in clp_dict[sec1title]:
+            clp_dict[sec1title][sec2title] = 0
+    # ... make sure every valid_dict section title is in clpdict
+    for sec1title, secdict in valid_dict.items():
+        assert isinstance(secdict, dict)
+        assert sec1title in clp_dict
+        for sec2title in secdict:
+            assert sec2title in clp_dict[sec1title]
+    # check validity of parameter section titles in docs/uguide.htmx skeleton
+    path = os.path.join(tests_path, '..', '..', 'docs', 'guide',
+                        'policy_params.md')
+    with open(path, 'r', encoding='utf-8') as md_file:
+        md_text = md_file.read()
+    md_dict = generate_section_dictionary(md_text)
+    # ... make sure every md_dict section title is in valid_dict
+    for sec1title,secdict in md_dict.items():
+        assert isinstance(secdict, dict)
+        assert sec1title in valid_dict
+        for sec2title in secdict:
+            assert sec2title in valid_dict[sec1title]
+    # ... make sure every valid_dict section title is in md_dict
+    for sec1title, secdict in valid_dict.items():
+        assert isinstance(secdict, dict)
+        assert sec1title in md_dict
+        for sec2title in secdict:
+            assert sec2title in md_dict[sec1title]
+
+
+def test_description_punctuation(tests_path):
+    """
+    Check that each description ends in a period.
+    """
+    # read JSON file into a dictionary
+    path = os.path.join(tests_path, '..', 'policy_current_law.json')
+    with open(path, 'r', encoding='utf-8') as jsonfile:
+        dct = json.load(jsonfile)
+        dct.pop("schema", None)
+    all_desc_ok = True
+    for param in dct.keys():
+        if not dct[param]['description'].endswith('.'):
+            all_desc_ok = False
+            print('param,description=',
+                  str(param),
+                  dct[param]['description'])
+    assert all_desc_ok
+
+
+def test_get_index_rate():
+    """
+    Test Parameters.get_index_rate.
+    """
+    pol = Policy()
+    wgrates = pol.get_index_rate('SS_Earnings_c', 2017)
+    pirates = pol.get_index_rate('II_em', 2017)
+    assert isinstance(wgrates, np.float64)
+    assert wgrates == pol.wage_growth_rates(2017)
+    assert pirates == pol.inflation_rates(2017)
+    assert isinstance(pirates, np.float64)
+    assert pol.inflation_rates() == pol._inflation_rates
+    assert pol.wage_growth_rates() == pol._wage_growth_rates
+
+
+def test_reform_with_bad_ctc_levels():
+    """
+    Implement a reform with _ACTC > _CTC_c values.
+    """
+    pol = Policy()
+    child_credit_reform = {
+        'CTC_c': {2020: 2200},
+        'ACTC_c': {2020: 2500}
+    }
+    with pytest.raises(pt.ValidationError):
+        pol.implement_reform(child_credit_reform)
+
+
+def test_reform_with_removed_parameter(monkeypatch):
+    """
+    Try to use removed parameter in a reform.
+    """
+    policy1 = Policy()
+    reform1 = {'FilerCredit_c': {2020: 1000}}
+    with pytest.raises(pt.ValidationError):
+        policy1.implement_reform(reform1)
+    policy2 = Policy()
+    reform2 = {'FilerCredit_c-indexed': {2020: True}}
+    with pytest.raises(pt.ValidationError):
+        policy2.implement_reform(reform2)
+
+    redefined_msg = {"some_redefined": "some_redefined was redefined."}
+    monkeypatch.setattr(Policy, "REDEFINED_PARAMS", redefined_msg)
+
+    pol = Policy()
+    with pytest.raises(pt.ValidationError):
+        pol.implement_reform({"some_redefined": "hello world"})
+
+
+def test_reform_with_out_of_range_error():
+    """
+    Try to use out-of-range values versus other parameter values in a reform.
+    """
+    pol = Policy()
+    reform = {'SS_thd85': {2020: [20000, 20000, 20000, 20000, 20000]}}
+    pol.implement_reform(reform, raise_errors=False)
+    assert pol.parameter_errors
+
+
+def test_reform_with_warning():
+    """
+    Try to use warned out-of-range parameter value in reform.
+    """
+    exp_warnings = {
+        'ID_Medical_frt': [
+            'ID_Medical_frt[year=2020] 0.05 < min 0.075 '
+        ]
+    }
+    pol = Policy()
+    reform = {'ID_Medical_frt': {2020: 0.05}}
+
+    pol.implement_reform(reform, print_warnings=True)
+    assert pol.warnings == exp_warnings
+    pol.set_state(year=2020)
+    assert pol.ID_Medical_frt == np.array([0.05])
+
+    pol.implement_reform(reform, print_warnings=False)
+    assert pol.warnings == {}
+    pol.set_state(year=2020)
+    assert pol.ID_Medical_frt == np.array([0.05])
+
+
+def test_reform_with_scalar_vector_errors():
+    """
+    Test catching scalar-vector confusion.
+    """
+    policy1 = Policy()
+    reform1 = {'SS_thd85': {2020: 30000}}
+    with pytest.raises(pt.ValidationError):
+        policy1.implement_reform(reform1)
+
+    policy2 = Policy()
+    reform2 = {'ID_Medical_frt': {2020: [0.08]}}
+    with pytest.raises(pt.ValidationError):
+        policy2.implement_reform(reform2)
+
+    policy3 = Policy()
+    reform3 = {'ID_Medical_frt': [{"year": 2020, "value": [0.08]}]}
+    with pytest.raises(pt.ValidationError):
+        policy3.adjust(reform3)
+
+    # Check that error is thrown if there are extra elements in array.
+    policy4 = Policy()
+    ref4 = {"II_brk1": {2020: [9700, 19400, 9700, 13850, 19400, 19400]}}
+    with pytest.raises(pt.ValidationError):
+        policy4.implement_reform(ref4)
+
+    policy5 = Policy()
+    ref5 = {"II_rt1": {2029: [.2, .3]}}
+    with pytest.raises(pt.ValidationError):
+        policy5.implement_reform(ref5)
+
+
+def test_index_offset_reform():
+    """
+    Test a reform that includes both a change in parameter_indexing_CPI_offset
+    and a change in a variable's indexed status in the same year.
+    """
+    # create policy0 to extract inflation rates before any
+    # parameter_indexing_CPI_offset
+    policy0 = Policy()
+    policy0.implement_reform({'parameter_indexing_CPI_offset': {2017: 0}})
+    cpiu_rates = policy0.inflation_rates()
+
+    reform1 = {'CTC_c-indexed': {2020: True}}
+    policy1 = Policy()
+    policy1.implement_reform(reform1)
+    offset = -0.005
+    reform2 = {'CTC_c-indexed': {2020: True},
+               'parameter_indexing_CPI_offset': {2020: offset}}
+    policy2 = Policy()
+    policy2.implement_reform(reform2)  # caused T-C crash before PR#2364
+    # extract from policy1 and policy2 the parameter values of CTC_c
+    pvalue1 = {}
+    pvalue2 = {}
+    for cyr in [2019, 2020, 2021]:
+        policy1.set_year(cyr)
+        pvalue1[cyr] = policy1.CTC_c[0]
+        policy2.set_year(cyr)
+        pvalue2[cyr] = policy2.CTC_c[0]
+    # check that pvalue1 and pvalue2 dictionaries contain the expected values
+    assert pvalue2[2019] == pvalue1[2019]
+    assert pvalue2[2020] == pvalue1[2020]
+    assert pvalue2[2020] == pvalue2[2019]
+    # ... indexing of CTC_c begins shows up first in 2021 parameter values
+    assert pvalue1[2021] > pvalue1[2020]
+    assert pvalue2[2021] > pvalue2[2020]
+    # ... calculate expected pvalue2[2021] from inflation rates and offset
+    syear = Policy.JSON_START_YEAR
+    expindexrate = cpiu_rates[2020 - syear] + offset
+    expvalue = round(pvalue2[2020] * (1. + expindexrate), 2)
+    # ... compare expected value with actual value of pvalue2 for 2021
+    assert np.allclose([expvalue], [pvalue2[2021]])
+
+
+def test_cpi_offset_affect_on_prior_years():
+    """
+    Test that parameter_indexing_CPI_offset does not have affect
+    on inflation rates in earlier years.
+    """
+    reform1 = {'parameter_indexing_CPI_offset': {2022: 0}}
+    reform2 = {'parameter_indexing_CPI_offset': {2022: -0.005}}
+    p1 = Policy()
+    p2 = Policy()
+    p1.implement_reform(reform1)
+    p2.implement_reform(reform2)
+
+    start_year = p1.start_year
+    p1_rates = np.array(p1.inflation_rates())
+    p2_rates = np.array(p2.inflation_rates())
+
+    # Inflation rates prior to 2022 are the same.
+    np.testing.assert_allclose(
+        p1_rates[:2022 - start_year],
+        p2_rates[:2022 - start_year]
+    )
+
+    # Inflation rate in 2022 was updated.
+    np.testing.assert_allclose(
+        p1_rates[2022 - start_year],
+        p2_rates[2022 - start_year] - (-0.005)
+    )
+
+
+def test_cpi_offset_on_reverting_params():
+    """
+    Test that params that revert to their pre-TCJA values
+    in 2026 revert if a parameter_indexing_CPI_offset is specified.
+    """
+    reform0 = {'parameter_indexing_CPI_offset': {2020: -0.001}}
+    reform1 = {'STD': {2017: [6350, 12700, 6350, 9350, 12700]},
+               'parameter_indexing_CPI_offset': {2020: -0.001}}
+    reform2 = {'STD': {2020: [10000, 20000, 10000, 10000, 20000]},
+               'parameter_indexing_CPI_offset': {2020: -0.001}}
+
+    p0 = Policy()
+    p1 = Policy()
+    p2 = Policy()
+    p0.implement_reform(reform0)
+    p1.implement_reform(reform1)
+    p2.implement_reform(reform2)
+
+    ryear = 2026
+    syear = Policy.JSON_START_YEAR
+
+    # STD was reverted in 2026
+    # atol=0.5 because ppp.py rounds params to nearest int
+    assert np.allclose(
+        p0._STD[ryear - syear],
+        p1._STD[ryear - syear], atol=0.5)
+
+    # STD was not reverted in 2026 if included in revision
+    assert not np.allclose(
+        p1._STD[ryear - syear],
+        p2._STD[ryear - syear], atol=0.5)
+
+
+def test_raise_errors_regression():
+    """
+    This tests that raise_errors prevents the error from being thrown. The
+    correct behavior is to exit the `adjust` function and store the errors.
+    """
+    ref = {
+        "II_brk7-indexed": [{"value": True}],
+        "II_brk6": [{"value": 316700, "MARS": "single", "year": 2020}],
+        "II_brk7": [{"value": 445400, "MARS": "single", "year": 2020}],
+
+    }
+    pol = Policy()
+    pol.adjust(ref, raise_errors=False)
+    assert pol.errors
+
+
+def test_simple_adj():
+    """
+    Test updating a 2D parameter that is indexed to inflation.
+    """
+    pol1 = Policy()
+    pol1.implement_reform(
+        {
+            "EITC_c": {
+                2020: [10000, 10001, 10002, 10003],
+                2023: [20000, 20001, 20002, 20003],
+            }
+        }
+    )
+    pol2 = Policy()
+    pol2.adjust(
+        {
+            "EITC_c": [
+                {"year": 2020, "EIC": "0kids", "value": 10000},
+                {"year": 2020, "EIC": "1kid", "value": 10001},
+                {"year": 2020, "EIC": "2kids", "value": 10002},
+                {"year": 2020, "EIC": "3+kids", "value": 10003},
+                {"year": 2023, "EIC": "0kids", "value": 20000},
+                {"year": 2023, "EIC": "1kid", "value": 20001},
+                {"year": 2023, "EIC": "2kids", "value": 20002},
+                {"year": 2023, "EIC": "3+kids", "value": 20003},
+            ]
+        }
+    )
+    cmp_policy_objs(pol1, pol2)
+
+    pol0 = Policy()
+    pol0.set_year(2019)
+    pol2.set_year(2019)
+
+    assert np.allclose(pol0.EITC_c, pol2.EITC_c)
+
+    pol2.set_state(year=[2020, 2021, 2022, 2023, 2024])
+    val2020 = np.array([[10000, 10001, 10002, 10003]])
+    val2023 = np.array([[20000, 20001, 20002, 20003]])
+
+    exp = np.vstack([
+        val2020,
+        val2020 * (1 + pol2.inflation_rates(year=2020)),
+        (
+            val2020 * (1 + pol2.inflation_rates(year=2020))
+        ).round(2) * (1 + pol2.inflation_rates(year=2021)),
+        val2023,
+        val2023 * (1 + pol2.inflation_rates(year=2023)),
+    ]).round(2)
+    np.testing.assert_allclose(pol2.EITC_c, exp)
+
+
+def test_adj_without_index_1():
+    """
+    Test update indexed parameter after turning off its indexed status.
+    """
+    pol1 = Policy()
+    pol1.implement_reform(
+        {
+            "EITC_c": {
+                2020: [10000, 10001, 10002, 10003],
+                2023: [20000, 20001, 20002, 20003],
+            },
+            "EITC_c-indexed": {2019: False},
+        }
+    )
+    pol2 = Policy()
+    pol2.adjust(
+        {
+            "EITC_c": [
+                {"year": 2020, "EIC": "0kids", "value": 10000},
+                {"year": 2020, "EIC": "1kid", "value": 10001},
+                {"year": 2020, "EIC": "2kids", "value": 10002},
+                {"year": 2020, "EIC": "3+kids", "value": 10003},
+                {"year": 2023, "EIC": "0kids", "value": 20000},
+                {"year": 2023, "EIC": "1kid", "value": 20001},
+                {"year": 2023, "EIC": "2kids", "value": 20002},
+                {"year": 2023, "EIC": "3+kids", "value": 20003},
+            ],
+            "EITC_c-indexed": [{"year": 2019, "value": False}],
+        }
+    )
+    cmp_policy_objs(pol1, pol2)
+
+    pol0 = Policy()
+    pol0.set_year(2019)
+    pol2.set_year(2019)
+
+    assert np.allclose(pol0.EITC_c, pol2.EITC_c)
+
+    pol2.set_state(year=[2020, 2021, 2022, 2023, 2024])
+
+    val2020 = np.array([[10000, 10001, 10002, 10003]])
+    val2023 = np.array([[20000, 20001, 20002, 20003]])
+
+    exp = np.vstack([
+        val2020,
+        val2020,
+        val2020,
+        val2023,
+        val2023,
+    ]).round(2)
+    np.testing.assert_allclose(pol2.EITC_c, exp)
+
+
+def test_adj_without_index_2():
+    """
+    Test updating an indexed parameter, making it unindexed,
+    and then adjusting it again.
+    """
+    pol1 = Policy()
+    pol1.implement_reform(
+        {
+            "EITC_c": {
+                2020: [10000, 10001, 10002, 10003],
+                2023: [20000, 20001, 20002, 20003],
+            },
+            "EITC_c-indexed": {2022: False},
+        }
+    )
+    pol2 = Policy()
+    pol2.adjust(
+        {
+            "EITC_c": [
+                {"year": 2020, "EIC": "0kids", "value": 10000},
+                {"year": 2020, "EIC": "1kid", "value": 10001},
+                {"year": 2020, "EIC": "2kids", "value": 10002},
+                {"year": 2020, "EIC": "3+kids", "value": 10003},
+                {"year": 2023, "EIC": "0kids", "value": 20000},
+                {"year": 2023, "EIC": "1kid", "value": 20001},
+                {"year": 2023, "EIC": "2kids", "value": 20002},
+                {"year": 2023, "EIC": "3+kids", "value": 20003},
+            ],
+            "EITC_c-indexed": [{"year": 2022, "value": False}],
+        }
+    )
+    cmp_policy_objs(pol1, pol2)
+
+    pol0 = Policy()
+    pol0.set_year(2019)
+    pol2.set_year(2019)
+
+    assert np.allclose(pol0.EITC_c, pol2.EITC_c)
+
+    pol2.set_state(year=[2020, 2021, 2022, 2023, 2024])
+
+    val2020 = np.array([[10000, 10001, 10002, 10003]])
+    val2023 = np.array([[20000, 20001, 20002, 20003]])
+
+    exp = np.vstack([
+        val2020,
+        val2020 * (1 + pol2.inflation_rates(year=2020)),
+        (
+            val2020 * (1 + pol2.inflation_rates(year=2020))
+        ).round(2) * (1 + pol2.inflation_rates(year=2021)),
+        val2023,
+        val2023,
+    ]).round(2)
+    np.testing.assert_allclose(pol2.EITC_c, exp)
+
+
+def test_activate_index():
+    """
+    Test changing a non-indexed parameter to an indexed parameter.
+    """
+    pol1 = Policy()
+    pol1.implement_reform({
+        "CTC_c": {2022: 2000},
+        "CTC_c-indexed": {2022: True}
+    })
+    pol2 = Policy()
+    pol2.adjust(
+        {
+            "CTC_c": [{"year": 2022, "value": 2000}],
+            "CTC_c-indexed": [{"year": 2022, "value": True}],
+        }
+    )
+    cmp_policy_objs(pol1, pol2)
+
+    pol0 = Policy()
+    pol0.set_year(year=2021)
+    pol2.set_state(year=[2021, 2022, 2023])
+    exp = np.array([
+        pol0.CTC_c[0],
+        2000,
+        2000 * (1 + pol2.inflation_rates(year=2022))
+    ]).round(2)
+
+    np.testing.assert_allclose(pol2.CTC_c, exp)
+
+
+def test_apply_cpi_offset():
+    """
+    Test applying the parameter_indexing_CPI_offset parameter
+    without any other parameters.
+    """
+    pol1 = Policy()
+    pol1.implement_reform(
+        {"parameter_indexing_CPI_offset": {2021: -0.001}}
+    )
+    pol2 = Policy()
+    pol2.adjust(
+        {"parameter_indexing_CPI_offset": [
+            {"year": 2021, "value": -0.001}
+        ]}
+    )
+    cmp_policy_objs(pol1, pol2)
+
+    pol0 = Policy()
+    pol0.implement_reform({"parameter_indexing_CPI_offset": {2021: 0}})
+
+    init_rates = pol0.inflation_rates()
+    new_rates = pol2.inflation_rates()
+
+    start_ix = 2021 - pol2.start_year
+
+    exp_rates = copy.deepcopy(new_rates)
+    exp_rates[start_ix:] -= pol2._parameter_indexing_CPI_offset[start_ix:]
+    np.testing.assert_allclose(init_rates, exp_rates)
+
+    # make sure values prior to 2021 were not affected.
+    cmp_policy_objs(pol0, pol2, year_range=range(pol2.start_year, 2021))
+
+    test_year = Policy.LAST_KNOWN_YEAR
+    pol2.set_state(year=[test_year, test_year + 1])
+    np.testing.assert_equal(
+        (pol2.EITC_c[1] / pol2.EITC_c[0] - 1).round(4),
+        (pol0.inflation_rates(year=test_year) + (-0.001)).round(4),
+    )
+
+
+def test_multiple_cpi_swaps():
+    """
+    Test changing a parameter's indexed status multiple times.
+    """
+    pol1 = Policy()
+    pol1.implement_reform(
+        {
+            "II_em": {2016: 6000, 2018: 7500, 2020: 9000},
+            "II_em-indexed": {2016: False, 2018: True},
+        }
+    )
+    pol2 = Policy()
+    pol2.adjust(
+        {
+            "II_em": [
+                {"year": 2016, "value": 6000},
+                {"year": 2018, "value": 7500},
+                {"year": 2020, "value": 9000},
+            ],
+            "II_em-indexed": [
+                {"year": 2016, "value": False},
+                {"year": 2018, "value": True},
+            ],
+        }
+    )
+    cmp_policy_objs(pol1, pol2)
+
+    # check inflation is not applied.
+    pol2.set_state(year=[2016, 2017])
+    np.testing.assert_equal(
+        pol2.II_em[0], pol2.II_em[1]
+    )
+
+    # check inflation rate is applied.
+    pol2.set_state(year=[2018, 2019])
+    np.testing.assert_equal(
+        (pol2.II_em[1] / pol2.II_em[0] - 1).round(4),
+        pol2.inflation_rates(year=2018),
+    )
+
+    # check inflation rate applied for rest of window.
+    window = list(range(2020, pol2.end_year + 1))
+    pol2.set_state(year=window)
+    np.testing.assert_equal(
+        (pol2.II_em[1:] / pol2.II_em[:-1] - 1).round(4),
+        [pol2.inflation_rates(year=year) for year in window[:-1]],
+    )
+
+
+def test_multiple_cpi_swaps2():
+    """
+    Test changing the indexed status of multiple parameters multiple times.
+    """
+    pol1 = Policy()
+    pol1.implement_reform(
+        {
+            "II_em": {2016: 6000, 2018: 7500, 2020: 9000},
+            "II_em-indexed": {2016: False, 2018: True},
+            "SS_Earnings_c": {2016: 300000, 2018: 500000},
+            "SS_Earnings_c-indexed": {2017: False, 2019: True},
+            "AMT_em-indexed": {2017: False, 2020: True},
+        }
+    )
+    pol2 = Policy()
+    pol2.adjust(
+        {
+            "SS_Earnings_c": [
+                {"year": 2016, "value": 300000},
+                {"year": 2018, "value": 500000},
+            ],
+            "SS_Earnings_c-indexed": [
+                    {"year": 2017, "value": False},
+                {"year": 2019, "value": True},
+            ],
+            "AMT_em-indexed": [
+                {"year": 2017, "value": False},
+                {"year": 2020, "value": True},
+            ],
+            "II_em": [
+                {"year": 2016, "value": 6000},
+                {"year": 2018, "value": 7500},
+                {"year": 2020, "value": 9000},
+            ],
+            "II_em-indexed": [
+                {"year": 2016, "value": False},
+                {"year": 2018, "value": True},
+            ],
+        }
+    )
+    cmp_policy_objs(pol1, pol2)
+
+    # Test SS_Earnings_c
+    # check inflation is still applied from 2016 to 2017.
+    pol2.set_state(year=[2016, 2017])
+    np.testing.assert_equal(
+        (pol2.SS_Earnings_c[1] / pol2.SS_Earnings_c[0] - 1).round(4),
+        pol2.wage_growth_rates(year=2016),
+    )
+
+    # check inflation rate is not applied after adjustment in 2018.
+    pol2.set_state(year=[2018, 2019])
+    np.testing.assert_equal(
+        pol2.SS_Earnings_c[0], pol2.SS_Earnings_c[1]
+    )
+
+    # check inflation rate applied for rest of window.
+    window = list(range(2019, pol2.end_year + 1))
+    pol2.set_state(year=window)
+    np.testing.assert_equal(
+        (pol2.SS_Earnings_c[1:] / pol2.SS_Earnings_c[:-1] - 1).round(4),
+        [pol2.wage_growth_rates(year=year) for year in window[:-1]],
+    )
+
+    # Test AMT
+    # Check values for 2017 through 2020 are equal.
+    pol2.set_state(year=[2017, 2018, 2019, 2020])
+    for i in (1, 2, 3):
+        np.testing.assert_equal(
+            pol2.AMT_em[0], pol2.AMT_em[i]
+        )
+
+    # check inflation rate applied for rest of window.
+    window = list(range(2020, pol2.end_year + 1))
+    pol2.set_state(year=window)
+    # repeat inflation rates accross matrix so they can be compared to the
+    # rates derived from AMT_em, a 5 * N matrix.
+    exp_rates = [pol2.inflation_rates(year=year) for year in window[:-1]]
+    exp_rates = np.tile([exp_rates], (5, 1)).transpose()
+    np.testing.assert_equal(
+        (pol2.AMT_em[1:] / pol2.AMT_em[:-1] - 1).round(4),
+        exp_rates,
+    )
+
+    # Test II_em
+    # check inflation is not applied.
+    pol2.set_state(year=[2016, 2017])
+    np.testing.assert_equal(
+        pol2.II_em[0], pol2.II_em[1]
+    )
+
+    # check inflation rate is applied.
+    pol2.set_state(year=[2018, 2019])
+    np.testing.assert_equal(
+        (pol2.II_em[1] / pol2.II_em[0] - 1).round(4),
+        pol2.inflation_rates(year=2018),
+    )
+
+    # check inflation rate applied for rest of window.
+    window = list(range(2020, pol2.end_year + 1))
+    pol2.set_state(year=window)
+    np.testing.assert_equal(
+        (pol2.II_em[1:] / pol2.II_em[:-1] - 1).round(4),
+        [pol2.inflation_rates(year=year) for year in window[:-1]],
+    )
+
+
+def test_adj_CPI_offset_and_index_status():
+    """
+    Test changing parameter_indexing_CPI_offset and another
+    parameter simultaneously.
+    """
+    pol1 = Policy()
+    pol1.implement_reform({
+        "CTC_c-indexed": {2020: True},
+        "parameter_indexing_CPI_offset": {2020: -0.005}},
+    )
+    pol2 = Policy()
+    pol2.adjust(
+        {
+            "parameter_indexing_CPI_offset":
+            [{"year": 2020, "value": -0.005}],
+            "CTC_c-indexed": [{"year": 2020, "value": True}],
+        }
+    )
+    cmp_policy_objs(pol1, pol2)
+
+    # Check no difference prior to 2020
+    pol0 = Policy()
+    pol0.implement_reform({"parameter_indexing_CPI_offset": {2020: 0}})
+    cmp_policy_objs(
+        pol0,
+        pol2,
+        year_range=range(pol2.start_year, 2020 + 1),
+        exclude=["parameter_indexing_CPI_offset"]
+    )
+
+    pol2.set_state(year=[2021, 2022])
+    np.testing.assert_equal(
+        (pol2.CTC_c[1] / pol2.CTC_c[0] - 1).round(4),
+        round(pol0.inflation_rates(year=2021) + (-0.005), 4),
+    )
+
+
+def test_adj_related_parameters_and_index_status():
+    """
+    Test changing two related parameters simulataneously and
+    one of their indexed statuses.
+    """
+    pol = Policy()
+    pol.adjust(
+        {
+            "II_brk7-indexed": [{"year": 2020, "value": True}],
+            # Update II_brk5 in 2026 to make reform valid after reset.
+            "II_brk5": [{"value": 330000, "MARS": "single", "year": 2026}],
+            "II_brk6": [{"value": 316700, "MARS": "single", "year": 2020}],
+            "II_brk7": [{"value": 445400, "MARS": "single", "year": 2020}],
+        }
+    )
+
+    # Check no difference prior to 2020
+    pol0 = Policy()
+    cmp_policy_objs(
+        pol0,
+        pol,
+        year_range=range(pol.start_year, 2019 + 1),
+    )
+
+    res = (
+        (pol.sel["II_brk6"]["MARS"] == "single")
+        & (pol.sel["II_brk6"]["year"] == 2020)
+    )
+    assert res.isel[0]["value"] == [316700]
+    res = (
+        (pol.sel["II_brk7"]["MARS"] == "single")
+        & (pol.sel["II_brk7"]["year"] == 2020)
+    )
+    assert res.isel[0]["value"] == [445400]
+
+    II_brk7 = pol.to_array("II_brk7", year=[2021, 2022])
+    II_brk7_single = II_brk7[:, 0]
+    np.testing.assert_equal(
+        (II_brk7_single[1] / II_brk7_single[0] - 1).round(4),
+        pol.inflation_rates(year=2021),
+    )
+
+
+def test_indexed_status_parsing():
+    """
+    Test parsing.
+    """
+    pol1 = Policy()
+    pol1.implement_reform({"EITC_c-indexed": {pol1.start_year: False}})
+    pol2 = Policy()
+    pol2.adjust({"EITC_c-indexed": False})
+    cmp_policy_objs(pol1, pol2)
+
+    with pytest.raises(pt.ValidationError):
+        pol2.adjust({"EITC_c-indexed": 123})
+
+
+def test_cpi_offset_does_not_affect_wage_indexed_params():
+    """
+    Test adjusting parameter_indexing_CPI_offset does not affect unknown
+    values of wage indexed parameters like SS_Earnings_c.
+    """
+    base_reform = {
+        "parameter_indexing_CPI_offset": {2021: -0.001},
+        "SS_Earnings_c": {2024: 300000},
+    }
+    pol0 = Policy()
+    pol0.implement_reform(base_reform)
+    pol1 = Policy()
+    pol1.implement_reform(base_reform)
+    pol1.implement_reform(dict(base_reform, SS_Earnings_c={2025: 500000}))
+
+    exp_before_2025 = pol0.to_array(
+        "SS_Earnings_c", year=list(range(2021, 2024 + 1))
+    )
+    act_before_2025 = pol1.to_array(
+        "SS_Earnings_c", year=list(range(2021, 2024 + 1))
+    )
+    np.testing.assert_equal(act_before_2025, exp_before_2025)
+
+
+def test_two_sets_of_tax_brackets():
+    """
+    Test that II_brk? and PT_brk? values are the same under current law.
+    """
+    pol = Policy()
+    brackets = range(1, 7+1)
+    years = range(Policy.JSON_START_YEAR, Policy.LAST_KNOWN_YEAR+1)
+    emsg = ''
+    for year in years:
+        pol.set_year(year)
+        pdata = dict(pol.items())
+        for bnum in brackets:
+            ii_val = pdata[f'II_brk{bnum}']
+            pt_val = pdata[f'PT_brk{bnum}']
+            if not np.allclose(ii_val, pt_val):
+                emsg += f'II_brk{bnum} != PT_brk{bnum} for year {year}\n'
+                emsg += f'  II_brk{bnum} is {ii_val}\n'
+                emsg += f'  PT_brk{bnum} is {pt_val}\n'
+    if emsg:
+        raise ValueError(emsg)