Table 5.

The effect of real interest rates on non-durable goods spending.

Wave
First follow-upSecond follow-upCombined follow-up
Panel A: Nielsen food
Posterior |$i - E\pi$|1.380.210.88
(1.17)(1.29)(1.02)
[|$-$|0.91, 3.67][|$-$|2.32, 2.74][|$-$|1.11, 2.88]
Observations11,69711,64311,578
F-statistic22.2122.5323.68
Panel B: Nielsen total
Posterior |$i - E\pi$|2.93***|$-$|0.471.46
(1.10)(1.27)(1.01)
[0.78, 5.07][|$-$|2.96, 2.02][|$-$|0.51, 3.43]
Observations11,65311,57411,527
F-statistic22.5022.5823.90
Panel C: survey food
Posterior |$i - E\pi$||$-$|0.32|$-$|0.090.67
(2.28)(3.13)(2.27)
[|$-$|4.79, 4.15][|$-$|6.73, 6.04][|$-$|3.78, 5.12]
Observations10,7515,4184,588
F-statistic14.596.7346.792
Panel D: survey non-durable
Posterior |$i - E\pi$|1.93|$-$|3.60|$-$|2.27
(1.95)(2.40)(2.18)
[|$-$|1.89, 5.76][|$-$|8.32, 1.11][|$-$|6.55, 2.01]
Observations10,7985,3724,555
F-statistic17.437.9785.972
Panel E: survey non-durable + debt payments
Posterior |$i - E\pi$|2.22|$-$|0.841.93
(1.88)(2.30)(2.34)
[|$-$|1.46, 5.91][-5.36, 3.67][|$-$|2.56, 6.53]
Observations10,8145,3674,527
F-statistic17.838.2346.954
Wave
First follow-upSecond follow-upCombined follow-up
Panel A: Nielsen food
Posterior |$i - E\pi$|1.380.210.88
(1.17)(1.29)(1.02)
[|$-$|0.91, 3.67][|$-$|2.32, 2.74][|$-$|1.11, 2.88]
Observations11,69711,64311,578
F-statistic22.2122.5323.68
Panel B: Nielsen total
Posterior |$i - E\pi$|2.93***|$-$|0.471.46
(1.10)(1.27)(1.01)
[0.78, 5.07][|$-$|2.96, 2.02][|$-$|0.51, 3.43]
Observations11,65311,57411,527
F-statistic22.5022.5823.90
Panel C: survey food
Posterior |$i - E\pi$||$-$|0.32|$-$|0.090.67
(2.28)(3.13)(2.27)
[|$-$|4.79, 4.15][|$-$|6.73, 6.04][|$-$|3.78, 5.12]
Observations10,7515,4184,588
F-statistic14.596.7346.792
Panel D: survey non-durable
Posterior |$i - E\pi$|1.93|$-$|3.60|$-$|2.27
(1.95)(2.40)(2.18)
[|$-$|1.89, 5.76][|$-$|8.32, 1.11][|$-$|6.55, 2.01]
Observations10,7985,3724,555
F-statistic17.437.9785.972
Panel E: survey non-durable + debt payments
Posterior |$i - E\pi$|2.22|$-$|0.841.93
(1.88)(2.30)(2.34)
[|$-$|1.46, 5.91][-5.36, 3.67][|$-$|2.56, 6.53]
Observations10,8145,3674,527
F-statistic17.838.2346.954

Notes: The table reports IV estimates of specification (2). The first stage is given by specification (1). The dependent variable is 100 |$\times$| log(monthly spending) over the last 3 months. Log(monthly spending) is trimmed at bottom and top 1%. Controls as in Table 4 are included but not reported. Panels A–E use different measures of spending. Panels A and B use scanner data from Nielsen. Panels C–E use self-reported spending data from the surveys. All regressions use sampling weights. Robust standard errors are in parentheses. 90% confidence intervals robust to weak IV are reported in square parentheses. “1|$^{\rm st}$| stage F-statistic” reports the F-statistic for the first-stage regression. The treatment of outliers and influential observations is described in Online Appendix C of Coibion et al. (2019). 90% confidence interval robust to weak IV is reported in square parentheses. *, **, and *** denote statistical significance at 10%, 5%, and 1% levels, respectively.

Table 5.

The effect of real interest rates on non-durable goods spending.

Wave
First follow-upSecond follow-upCombined follow-up
Panel A: Nielsen food
Posterior |$i - E\pi$|1.380.210.88
(1.17)(1.29)(1.02)
[|$-$|0.91, 3.67][|$-$|2.32, 2.74][|$-$|1.11, 2.88]
Observations11,69711,64311,578
F-statistic22.2122.5323.68
Panel B: Nielsen total
Posterior |$i - E\pi$|2.93***|$-$|0.471.46
(1.10)(1.27)(1.01)
[0.78, 5.07][|$-$|2.96, 2.02][|$-$|0.51, 3.43]
Observations11,65311,57411,527
F-statistic22.5022.5823.90
Panel C: survey food
Posterior |$i - E\pi$||$-$|0.32|$-$|0.090.67
(2.28)(3.13)(2.27)
[|$-$|4.79, 4.15][|$-$|6.73, 6.04][|$-$|3.78, 5.12]
Observations10,7515,4184,588
F-statistic14.596.7346.792
Panel D: survey non-durable
Posterior |$i - E\pi$|1.93|$-$|3.60|$-$|2.27
(1.95)(2.40)(2.18)
[|$-$|1.89, 5.76][|$-$|8.32, 1.11][|$-$|6.55, 2.01]
Observations10,7985,3724,555
F-statistic17.437.9785.972
Panel E: survey non-durable + debt payments
Posterior |$i - E\pi$|2.22|$-$|0.841.93
(1.88)(2.30)(2.34)
[|$-$|1.46, 5.91][-5.36, 3.67][|$-$|2.56, 6.53]
Observations10,8145,3674,527
F-statistic17.838.2346.954
Wave
First follow-upSecond follow-upCombined follow-up
Panel A: Nielsen food
Posterior |$i - E\pi$|1.380.210.88
(1.17)(1.29)(1.02)
[|$-$|0.91, 3.67][|$-$|2.32, 2.74][|$-$|1.11, 2.88]
Observations11,69711,64311,578
F-statistic22.2122.5323.68
Panel B: Nielsen total
Posterior |$i - E\pi$|2.93***|$-$|0.471.46
(1.10)(1.27)(1.01)
[0.78, 5.07][|$-$|2.96, 2.02][|$-$|0.51, 3.43]
Observations11,65311,57411,527
F-statistic22.5022.5823.90
Panel C: survey food
Posterior |$i - E\pi$||$-$|0.32|$-$|0.090.67
(2.28)(3.13)(2.27)
[|$-$|4.79, 4.15][|$-$|6.73, 6.04][|$-$|3.78, 5.12]
Observations10,7515,4184,588
F-statistic14.596.7346.792
Panel D: survey non-durable
Posterior |$i - E\pi$|1.93|$-$|3.60|$-$|2.27
(1.95)(2.40)(2.18)
[|$-$|1.89, 5.76][|$-$|8.32, 1.11][|$-$|6.55, 2.01]
Observations10,7985,3724,555
F-statistic17.437.9785.972
Panel E: survey non-durable + debt payments
Posterior |$i - E\pi$|2.22|$-$|0.841.93
(1.88)(2.30)(2.34)
[|$-$|1.46, 5.91][-5.36, 3.67][|$-$|2.56, 6.53]
Observations10,8145,3674,527
F-statistic17.838.2346.954

Notes: The table reports IV estimates of specification (2). The first stage is given by specification (1). The dependent variable is 100 |$\times$| log(monthly spending) over the last 3 months. Log(monthly spending) is trimmed at bottom and top 1%. Controls as in Table 4 are included but not reported. Panels A–E use different measures of spending. Panels A and B use scanner data from Nielsen. Panels C–E use self-reported spending data from the surveys. All regressions use sampling weights. Robust standard errors are in parentheses. 90% confidence intervals robust to weak IV are reported in square parentheses. “1|$^{\rm st}$| stage F-statistic” reports the F-statistic for the first-stage regression. The treatment of outliers and influential observations is described in Online Appendix C of Coibion et al. (2019). 90% confidence interval robust to weak IV is reported in square parentheses. *, **, and *** denote statistical significance at 10%, 5%, and 1% levels, respectively.

Close
This Feature Is Available To Subscribers Only

Sign In or Create an Account

Close

This PDF is available to Subscribers Only

View Article Abstract & Purchase Options

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

Close