Calibration of Household Preference Parameters#

Behavioral Assumptions#

Elasticity of labor supply#

As discussed in the OG-Core household theory documentation, we use the elliptical disutility of labor function developed by [Evans and Phillips, 2017]. We then fit the parameters of the elliptical utility function to match the marginal disutility from a constant Frisch elasticity function. OG-ZAF users enter the constant Frisch elasticity as a parameter. [Peterman, 2016] finds a range of Frisch elasticities estimated from microeconomic and macroeconomic data. These range from 0 to 4. Peterman makes the case that in lifecycle models without an extensive margin for employment the Frisch elasticity should be higher. For OG-ZAF we take a default value of 0.4 from [Altonji, 1986].

Intertemporal elasticity of substitution#

The default value for the intertemporal elasticity of substitution, \(\sigma\), is taken from [Attanasio et al., 1999]. We set \(\sigma=1.5\).

Rate of time preference#

The rate of time preference is calibrated, not borrowed. With the capital share, depreciation and tax rates set from data, \(\beta\) moves the return on capital and so the capital-output ratio. We set \(\beta = 0.8706\) (annual, the same for every lifetime-income group), the value at which the steady-state private capital-output ratio equals 2.80 (see the firms chapter). This is lower than the 0.96 common in models of high-saving economies, and it matches South Africa’s low domestic saving: gross national saving in 2025 was about 13.4% of GDP (gross capital formation of 13.9% less a current account deficit of 0.5%, SARB Quarterly Bulletin Tables S-116 and S-159). The implied real return on capital is 11.0%, against PWT’s measured 12.0%.