11 tithing & income

11_income_myth_2.py

import pandas as pd
import numpy as np
import survey_tools as st
import scipy as sp

# read in the PRLS
data_prls = pd.read_spss('../data/rls/rls2024.sav')

data_prls['inc'] = data_prls.INC_SDT1.replace({
    'Less than $30,000': 'Less than 30k',
    '$30,000 to less than $40,000': '30k to 50k',
    '$40,000 to less than $50,000': '30k to 50k',
    '$50,000 to less than $75,000': '50k to 75k',
    '$75,000 to less than $100,000': '75k to 100k',
    '$100,000 to less than $125,000': '100k to 150k',
    '$125,000 to less than $150,000': '100k to 150k',
    '$150,000 or more': 'More than 150k',
    "DON'T KNOW/REFUSED": np.nan,
})

data_prls['inc'].cat.set_categories(
    ['Less than 30k', '30k to 50k', '50k to 75k', '75k to 100k', '100k to 150k', 'More than 150k'],
    inplace=True,
    ordered=True
)

data_prls['attend'] = np.where(data_prls.ATTNDPERRLS.isin(['More than once a week', 'Once a week']), "Weekly", "Less than weekly")

final_df = pd.DataFrame({
    'income_cat' : data_prls.inc.cat.categories,
    'overall_prls': st.tabs(data_prls,'inc', display='column', wts='WEIGHT').to_list(),
    'mormon_prls': st.tabs(data_prls.query('RELTRAD == "Church of Jesus Christ of Latter-day Saints (Mormon)"'),'inc', display='column').to_list(),
    'mormon_active_prls': st.tabs(data_prls.query('RELTRAD == "Church of Jesus Christ of Latter-day Saints (Mormon)" & attend == "Weekly"'),'inc', display='column').to_list(),
    'mormon_lessactive_prls': st.tabs(data_prls.query('RELTRAD == "Church of Jesus Christ of Latter-day Saints (Mormon)" & attend == "Less than weekly"'),'inc', display='column').to_list(),
    'tithe_payers_2023': [11.7, 9.6, 18.8, 18.3, 23.2, 18.4], #hardcoded from CFLDS
    'non_tithe_payers_2023': [14.5, 16.7, 24, 12.8, 19, 13] #hardcoded from CFLDS
})

final_df = final_df.set_index('income_cat')
final_df.T.loc[['overall_prls', 'mormon_prls'],:]

# let's do some tests...
# let's test Mormons vs. Overall
sp.stats.ttest_ind(
    data_prls[~data_prls.inc.isna()].inc.cat.codes.to_list(),
    data_prls.query('RELTRAD == "Church of Jesus Christ of Latter-day Saints (Mormon)" & inc.notna()').inc.cat.codes.to_list(),
)
sp.stats.mannwhitneyu(
    data_prls[~data_prls.inc.isna()].inc.cat.codes.to_list(),
    data_prls.query('RELTRAD == "Church of Jesus Christ of Latter-day Saints (Mormon)" & inc.notna()').inc.cat.codes.to_list(),
    alternative='two-sided'
)
# mormons not statistically different from overall

# Now let's test less active vs. active Mormons
sp.stats.ttest_ind(
    data_prls.query('RELTRAD == "Church of Jesus Christ of Latter-day Saints (Mormon)" & attend == "Weekly" & inc.notna()').inc.cat.codes.to_list(),
    data_prls.query('RELTRAD == "Church of Jesus Christ of Latter-day Saints (Mormon)" & attend == "Less than weekly" & inc.notna()').inc.cat.codes.to_list(),
)
sp.stats.mannwhitneyu(
    data_prls.query('RELTRAD == "Church of Jesus Christ of Latter-day Saints (Mormon)" & attend == "Weekly" & inc.notna()').inc.cat.codes.to_list(),
    data_prls.query('RELTRAD == "Church of Jesus Christ of Latter-day Saints (Mormon)" & attend == "Less than weekly" & inc.notna()').inc.cat.codes.to_list(),
    alternative='greater' # i expect active Mormons to have higher income
)
# active Mormons have statistically significantly higher income than less active Mormons

# active Mormons vs. the overall population
sp.stats.ttest_ind(
    data_prls.query('RELTRAD == "Church of Jesus Christ of Latter-day Saints (Mormon)" & attend == "Weekly" & inc.notna()').inc.cat.codes.to_list(),
    data_prls[~data_prls.inc.isna()].inc.cat.codes.to_list(),
    alternative='greater' # i expect active Mormons to have higher income than the overall population
)
sp.stats.mannwhitneyu(
    data_prls.query('RELTRAD == "Church of Jesus Christ of Latter-day Saints (Mormon)" & attend == "Weekly" & inc.notna()').inc.cat.codes.to_list(),
    data_prls[~data_prls.inc.isna()].inc.cat.codes.to_list(),
    alternative='greater' # i expect active Mormons to have higher income than the overall population
)
# active Mormons have statistically significantly higher income than the overall population at 90% confidence

# less active Mormons vs. the overall population
sp.stats.ttest_ind(
    data_prls.query('RELTRAD == "Church of Jesus Christ of Latter-day Saints (Mormon)" & attend == "Less than weekly" & inc.notna()').inc.cat.codes.to_list(),
    data_prls[~data_prls.inc.isna()].inc.cat.codes.to_list(),
    alternative='less'
)
sp.stats.mannwhitneyu(
    data_prls.query('RELTRAD == "Church of Jesus Christ of Latter-day Saints (Mormon)" & attend == "Less than weekly" & inc.notna()').inc.cat.codes.to_list(),
    data_prls[~data_prls.inc.isna()].inc.cat.codes.to_list(),
    alternative='less' # i expect less active Mormons to have higher income than the overall population
)
# active Mormons have statistically significantly lower income than the overall population