Table 5.
Falsification Tests: Regressions Using Cognition Outcomes
| Financial Literacy | Calculation | Number Series | Matrix Reasoning | Verbal Analogies | Auditory Memory | Visual Matching | |
|---|---|---|---|---|---|---|---|
| Mean dependent variable | −0.367 | −0.316 | −0.192 | 0.027 | −0.198 | 0.251 | 0.383 |
|
| |||||||
| Expected time to widowhood | −0.041** (0.019) | −0.001 (0.018) | −0.009 (0.017) | −0.036** (0.015) | −0.019 (0.017) | −0.031* (0.018) | −0.017 (0.018) |
| Expected duration of widowhood | 0.051 (0.039) | 0.043 (0.037) | 0.006 (0.034) | 0.044 (0.031) | 0.018 (0.036) | 0.004 (0.036) | 0.063* (0.037) |
| Woman’s education | 0.06 (0.041) | 0.144*** (0.040) | 0.187*** (0.036) | 0.094*** (0.033) | 0.144*** (0.038) | 0.047 (0.039) | 0.068* (0.039) |
| Husband’s education | −0.107*** (0.038) | −0.147*** (0.037) | −0.123*** (0.034) | −0.067** (0.031) | −0.069* (0.035) | −0.083** (0.035) | −0.085** (0.036) |
| Woman’s health | 0.058 (0.079) | 0.192** (0.076) | 0.039 (0.070) | 0.178*** (0.063) | −0.029 (0.073) | 0.116 (0.074) | 0.147* (0.075) |
| Husband’s health | −0.08 (0.075) | −0.026 (0.071) | 0.072 (0.066) | 0.012 (0.059) | 0.069 (0.069) | −0.072 (0.070) | 0.042 (0.070) |
| Wealth decile | −0.017 (0.028) | −0.049* (0.027) | −0.045* (0.025) | −0.039* (0.023) | −0.034 (0.026) | −0.018 (0.026) | −0.056** (0.027) |
| Constant | 0.435 (0.740) | −1.104 (0.710) | −1.145* (0.651) | −0.858 (0.591) | −1.154* (0.682) | 1.138 (0.692) | −0.288 (0.707) |
| R2 | 0.066 | 0.136 | 0.134 | 0.093 | 0.074 | 0.069 | 0.077 |
| F | 2.193 | 4.832 | 4.726 | 3.146 | 2.401 | 2.227 | 2.493 |
| Observations | 224 | 223 | 222 | 223 | 219 | 219 | 217 |
Notes:
p<0.1
p<0.05
p<0.01.
OLS regressions with dependent variables: wife - husband difference in normalized cognition scores as specified in the column labels. These falsification tests reproduce the specification in column (2) of Table 4. The cognition variables are detailed in Section IV.C. Because the cognition scores are normalized, the coefficients are directly comparable with those estimated using the wife-husband difference in financial literacy in column (1).
Data sources: CogEcon and CogUSA.