Table 4.
The simulated association of serious mental illness with reduced earnings at the individual level and societal level in each World Mental Health country
| Serious mental illness
|
The associations expressed as a proportion of mean national
earningsa |
The associations expressed in local currencyb |
||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Prevalence
|
Individual-levelc |
Societal-leveld |
Individual-level
|
Societal-level (in Billions)
|
||||||
| % | (s.e.) | Estimate | (s.e.) | Estimate | (s.e.) | Estimate | (s.e.) | Estimate | (s.e.) | |
| High-income | ||||||||||
| Belgium | 4.9 | (1.0) | 27.3 | (17.3) | 1.3 | (0.8) | 211 807 | (135 679) | 65.8 | (42.2) |
| France | 3.8 | (0.5) | 38.9* | (18.7) | 1.5* | (0.7) | 42 660* | (20 473) | 56.7* | (27.0) |
| Germany | 2.7 | (0.5) | 36.8 | (36.8) | 1.0 | (1.0) | 11 908 | (11 886) | 16.8 | (16.7) |
| Israel | 3.7 | (0.3) | 23.8* | (10.0) | 0.9* | (0.4) | 1516* | (622) | 0.2* | (0.0) |
| Italy | 1.3 | (0.2) | 1.5 | (7.9) | 0.0 | (0.1) | 409 686 | (2 107 357) | 194.1 | (998.7) |
| Japan | 1.2 | (0.4) | 19.1 | (30.0) | 0.2 | (0.4) | 611 752 | (959 881) | 5.4 | (8.7) |
| The Netherlands | 4.4 | (0.7) | 16.2* | (6.2) | 0.7* | (0.3) | 7404* | (2901) | 3.5* | (1.4) |
| New Zealand | 4.9 | (0.3) | 25.3 | (15.5) | 1.3 | (0.8) | 10 031 | (6275) | 1.2 | (1.3) |
| Spain | 1.9 | (0.3) | 18.0 | (15.2) | 0.3 | (0.3) | 353 356 | (294 691) | 177.2 | (147.6) |
| United States
|
6.5
|
(0.4)
|
25.5*
|
(6.5)
|
1.7*
|
(0.4)
|
8519*
|
(2100)
|
99.5*
|
(24.6)
|
| Low- and middle-income | ||||||||||
| Brazil | 9.3 | (0.7) | 1.5 | (2.5) | 0.1 | (0.2) | 17 | (30) | 0.0 | (0.0) |
| Bulgaria | 1.5 | (0.3) | 26.6 | (18.7) | 0.4 | (0.3) | 638 | (441) | 0.0 | (0.0) |
| Colombia | 4.1 | (0.4) | 20.1 | (13.0) | 0.8 | (0.5) | 1 051 625 | (675 762) | 1042.9 | (670.3) |
| India (Pondicherry) | 1.0 | (0.2) | 39.4 | (125.9) | 0.4 | (1.3) | 17 478 | (55 715) | 0.1* | (0.0) |
| Lebanon | 4.1 | (0.7) | 2.1 | (1.1) | 0.1* | (0.0) | 141* | (70) | 0.0 | (0.0) |
| Mexico | 2.3 | (0.2) | 5.9 | (4.6) | 0.1 | (0.1) | 2022 | (1622) | 2.4 | (1.5) |
| Nigeria | 0.5 | (0.2) | 34.5 | (56.5) | 0.2 | (0.3) | 23 599 | (38 745) | 7.5 | (12.4) |
| People’s Republic of China (Beijing, Shanghai) | 0.6 | (0.2) | 28.9 | (57.4) | 0.2 | (0.3) | 413 | (826) | 0.0 | (0.0) |
| Shenzhen | 0.9 | (0.3) | 3.6 | (10.9) | 0.0 | (0.1) | 1141 | (3413) | 0.0 | (0.0) |
| South Africa
|
3.3
|
(0.3)
|
18.8
|
(22.3)
|
0.6
|
(0.7)
|
4798
|
(5585)
|
3.9
|
(4.8)
|
| Pooled | ||||||||||
| High-income | 4.3 | (0.1) | 19.4* | (1.8) | 0.8* | (0.1) | ||||
| Low- and middle-income | 3.1 | (0.1) | 10.9* | (4.6) | 0.3* | (0.1) | ||||
a. Results are expressed here in terms of mean earnings, whereas they were expressed in terms of median earnings in Table 3. The median was used in estimating the models in Table 3 because this transformation was considered the one that makes most sense as the basis for constraining model coefficients to be constant across countries. The mean is used here, in comparison, because it is the natural metric for interpreting the substantive meaning of results. To clarify the interpretation: if 4.3% of respondents in high-income countries have serious mental illness and serious mental illness is associated with a 19.4% reduction in earnings, then this level of loss in this segment of the population represents 0.194×0.043 = 0.8% of all national earnings.
b. The local currencies are francs in Belgium, francs in France, marks in Germany, shekels in Israel, lira in Italy, yen in Japan, guilders in The Netherlands, dollars in New Zealand, pesetas in Spain, dollars in the USA, reals in Brazil, lev in Bulgaria, pesos in Colombia, rupees in India, pounds in Lebanon, pesos in Mexico, naira in Nigeria, yuan in People’s Republic of China, and rand in South Africa.
c. Estimates do not differ significantly across either high-income countries (χ2(9) = 8.1, P = 0.52) or low- and middle-income countries (χ2(9) = 5.5, P = 0.79) based on design-based Wald χ2-tests.
d. Estimates differ significantly across high-income countries (χ2(9) = 30.4, P<0.001) but not low- and middle-income countries (χ2(9) = 4.7, P = 0.86) based on design-based Wald χ2-tests.
*Significant at the 0.05 level, two-sided test.