Nuptiality
Marriage markets appear to have been integrated within regions but not between them. Variation between regions was more apparent for male marriage than for
fertility. Whereas 88.6% of men between the ages of 36 and 50 sui in the north were or had been married, only 83.3% in Daoyi and 84.1% in the other central farm systems had been married.
Once again, the south was in the middle. Variation within regions was far less pronounced than was the case with marital fertility, Dami aside. Proportions married in the central state farm systems were almost identical and varied little in the northern and southern systems. We are hesitant to draw any conclusions from a comparison of levels of male marriage because they may be affected by conditions outside the farm systems. For example, low proportions married in Daoyi may simply reflect the effects of proximity to a large city, Shenyang.Long-term stability in the proportions of men in late middle age who had ever married suggests that the overall supply of females changed little. According to Figure 16.4, between 1760 and 1910, the percentage of men aged 36—45 who had ever married hovered between 80 and 90. Since the major source of variation over time in the supply of marriageable females was female infanticide and neglect, the implication is that there was no secular trend in their incidence. To the extent that female infanticide and neglect was a response to economic stress, there is little evidence here of a long-term decline in living standards.
Inequality in male access to marriage, however, increased over time: according to Figure 16.4 the range of ages over which men first married widened substantially. In other words, a progressively largely proportion of families found it within their wherewithal to acquire brides for their sons while they were still young. Thus the proportion of men marrying at young ages increased steadily, even though the proportion of men who ever married by the time they reached late middle age remained
Figure 16.4 Percentage of men married at different ages in Liaoning
stable or perhaps even fell slightly.
The proportion of men aged 16—25 who had ever married doubled from one- quarter to one-half between 1760 and 1910.Event-history analysis confirms the increase in inequality in male access to marriage. Coefficients in Table 16.4 indicate that while those who would marry did so at younger and younger ages, those who had not married by the time they reached adulthood were steadily less likely to never marry. The chances that a man aged 6—15 would marry in the next three years increased in every region from 1774 and 1888. The smallest increase was in the central region, where chances rose by 0.38% a year, and highest in the south, where they rose by 1.33% a year. Over 100 years, in other words, the chances of marrying in the next three years increased by 46% in the north and 278% in the south. Coefficients for the age ranges 16—25 and 26—40, meanwhile, indicate that in every region the chances that a man who was still unmarried after age 15 would marry in the next three years declined.
Male marriage does not appear to have responded in a consistent fashion to current grain prices.17 Only in the central region between 1774 and 1834 was the
Table 16.4 Coefficients for year from the complementary log—log regression of marriage in the next three years for never-married males
| Model 1: 1744 | -1888 | Model 2: 1774 | -1834 | Model 3: 1834 | -88 | |
| Coefficient | p-value | Coefficient | p-valιue | Coefficient | p-valιue | |
| Age 6-15 | ||||||
| North | 0.0114 | 0.00 | 0.0124 | 0.03 | 0.0120 | 0.00 |
| Central | 0.0038 | 0.08 | 0.0310 | 0.00 | -0.0031 | 0.49 |
| Daoyi | 0.0099 | 0.00 | 0.0343 | 0.00 | 0.0204 | 0.00 |
| South | 0.0133 | 0.00 | 0.0113 | 0.10 | 0.0174 | 0.00 |
| N | 33,276 | 14,536 | 18,740 | |||
| Age 16-25 | ||||||
| North | -0.0060 | 0.00 | -0.0098 | 0.00 | -0.0056 | 0.03 |
| Central | -0.0036 | 0.01 | 0.0105 | 0.06 | -0.0034 | 0.36 |
| Daoyi | -0.0020 | 0.06 | -0.0061 | 0.19 | 0.0065 | 0.04 |
| South | -0.0013 | 0.20 | 0.0123 | 0.01 | -0.0007 | 0.79 |
| N | 17,277 | 8,416 | 8,861 | |||
| Age 26-40 | ||||||
| North | -0.0015 | 0.39 | -0.0156 | 0.67 | 0.0087 | 0.03 |
| Central | -0.0045 | 0.12 | 0.0050 | 0.43 | -0.0076 | 0.25 |
| Daoyi | -0.0086 | bgcolor=white>0.00-0.0066 | 0.17 | 0.0012 | 0.87 | |
| South | -0.0059 | 0.00 | 0.0664 | 0.02 | -0.0076 | 0.12 |
| N | 8,229 | 3,534 | 4,695 |
Notes: The models did not include main effects of year or low sorghum price, only the interactions between them and the four dichotomous indicators of region. To save space, coefficients for the region indicators and their interactions with the terms of the orthogonal age polynomial are omitted from this table.
We restricted analysis to observations where the immediately succeeding observation was also available.effect of low sorghum prices in the expected direction. A 10% increase in grain prices lowered the chances of marriage in the next three years by 10.9% for males age 6—15, and by 6.1% for males aged 16—25. In the other regions and time periods, increases in sorghum prices either had no effect, or else increased the chances of marriage. Previously, we have argued that such results may indicate that reductions during bad times in the financial capacity of households to secure brides for their sons may have been offset by increased desperation by households with daughters to reduce expenses or raise funds by marrying them off (Campbell and Lee 1998).
4.3