Key Decisions in the Lives of the Old Order Amish: Joining the Church and Migrating to Another Settlement
LAWRENCE P. GREKSA AND JILL E. KORBIN*
Abstract: The rates at which adult children migrate to other Amish settlements and decide to leave the Amish church were derived from the 1993 directory for the Old Order Amish living in the Geauga settlement in northeast Ohio. Out-migration rates were maintained at a fairly high level of 14-20% of all adult children born between 1928 and 1967, but the rate at which adult children left the Amish church decreased from about 30% to 5% over this time period. The majority of the men in the Geauga settlement are now wage laborers but, contrary to the pattern in the Lancaster settlement, the transition from farming to wage labor has not resulted in an increase in the rate at which adult children leave the church. The effects of whether the father is a religious leader, family size and parity on rates of out-migration and leaving the church are also examined. Finally, the implications of these findings for the future of the Amish church are discussed.
All Old Order Amish adults are socially constrained to make several conscious and public decisions with reasonably well understood consequences. First, they must decide whether or not they will join the Amish church. Second, if they do join the church, they must decide if they will live in their birth settlement or migrate to a new settlement. The fact that these two events are discussed together is not meant to imply that they are similar behaviors or are linked in any functional manner. In fact, the factors influencing the decision to join the church, as well as the social ramifications of this decision, are clearly substantially different and of a greater magnitude than those associated with the decision to migrate to another settlement. The rationale for considering these two very different events concurrently is that they both influence the demographic structure of a settlement.
The decision of whether or not to join the church, which occurs during late adolescence or early adulthood, is clearly the most important decision that an Amish adult will make. If adults decide not to join the church, they do so knowing that they are giving up the lifestyle of their childhood and that they will have less social contact and receive less social support from their family and the other members of their community of birth. Since the initial decision to join the church is seen by the community as a decision for life, the social consequences of later leaving the church include the imposition of excommunication and Meidung (shunning), which act to isolate individuals from their family and community to a much greater extent (often completely) than if they never join the church.
A second decision is whether they will remain in their birth settlement or migrate to another Amish settlement. This decision can be made at any time during adulthood, although it generally appears to occur during early and middle adulthood. Amish adults migrate for a variety of reasons but the two primary reasons are the desire for less expensive farm land and the desire to live in a congregation with either a more or less conservative Ordnung. Out-migration by necessity also results in some reduction in social contact (and thus reduced social support) with relatives and other members of the birth settlement. However, unlike the kind of reduced social support that occurs when someone leaves the church, this reduced support is simply a function of distance, not community-imposed social constraints. As a result, someone living far from the rest of their family will not receive the kind of daily contact and support that can occur in the birth settlement, but they can expect to receive substantial support in the event of an emergency. In addition, a family can always decide to migrate back to their home settlement. Individuals raised in a society that emphasizes community solidarity and mutual aid fully understand the consequences of both of these decisions.
Although it is well understood that these decisions, especially the decision of whether or not to join the Amish church, must be made by all adult Amish, there is only limited information in research literature on the manner in which Amish adults make these two decisions. This present report will describe the rates of these two decisions in the Geauga settlement in northeast Ohio and describe some of the factors that influence these rates, generally following the model provided by sociologist Thomas Meyers for the Elkhart-LaGrange settlement. 
The present report is primarily based on the directory for the Geauga settlement covering the period up to January 1, 1993. An earlier report used information derived from this directory to describe the fertility of the Geauga settlement. The present report utilizes data from the directory on the status of a couple’s adult children. Previous population studies based on Amish directories have found them to be about 98% complete.
Directory information on the status of children was used to classify all adult children (defined as 18 years and older) into one of three categories: (1) Amish and living in the Geauga settlement; (2) Amish but living in another settlement; or (3) definitely or possibly non-Amish. Individuals who left the Amish community before they joined the church and those who left after joining the church cannot be distinguished with the available information and are thus lumped together. About 71% of the non-Amish category includes individuals who can be classified as non-Amish with a high degree of confidence. This group includes (a) adult children whose parents specifically identified them as no longer being Amish; and (b) adult children who were not given a status by their parents and who are not listed in the Geauga settlement directory but who were given a local address. The remaining 29% of this category includes adult children who were given no status by their parents and who were not listed in the Geauga settlement directory. At least some of these individuals may be Amish but living in another settlement, making these estimates conservative. For simplicity, the process of becoming non-Amish is referred to as “leaving the church,” even though technically this description applies only to those who first join the church.
Sam Weaver, who migrated from Holmes County, Ohio in 1886, was reportedly the first Amish settler in Geauga County. He was soon followed by other migrants from Holmes County, as well as some from western Pennsylvania, Indiana and elsewhere. As the settlement grew, Amish families moved into nearby Trumbull, Ashtabula and Portage Counties. There are no good estimates of the size of the Geauga settlement during its early years, but it appears that the settlement initially grew fairly slowly. Based on the directories compiled by the settlement, the size of the settlement was 7546 individuals in 1428 families living in 47 congregations in 1988; 8333 individuals in 1608 families living in 54 congregations in 1993; and 9572 individuals in 1838 families living in 63 congregations in 1998. The rate of growth between 1988 and 1998 was 2.4%, with a doubling time of about 29 years.
For many years the Amish have been undergoing a well documented economic transition away from small, self-sufficient family farms and towards various kinds of wage labor. The magnitude of this transition in the Geauga settlement is portrayed in Table 1, which describes the frequency of occupational categories for the male heads of household based on both the 1973 directory (the earliest for the Geauga settlement) and the 1993 directory. Heads of household were considered to be farmers if they listed farmer or retired farmer as either their primary or secondary occupation.
FREQUENCY OF OCCUPATIONAL CATEGORIES IN THE GEAUGA SETTLEMENT IN 1973 AND 1993
N % N %
Farming 260 31.1 269 16.8
Shop 100 12.0 139 8.7
Factory and Lumber 249 29.8 541 33.7
Construction 154 18.4 423 26.4
Other 72 8.6 231 14.4
(2 = 75.5, p < .001
a Occupation missing for 45 families
b Occupation missing for 28 families
Shops tend to be located on a home property, so that shop owners, like farmers, work at home. Men working in a factory, lumber yard or lumber mill were grouped together, since each occupation generally involves working away from home in settings with at least some non-Amish co-workers. Construction workers also work away from home but they are in a separate category since most Amish men work in largely or completely Amish crews. Since many men do not clearly fall into any of these categories, their occupation was classified as “other.” Some of the men in the “other” category described their occupations in ways which made it impossible to determine the kind of work they perform (e.g., day worker, day laborer); others have occupations that do not clearly fit into any category (e.g., golf course employee, butcher); while others used labels that could apply to either shop workers or factory workers (e.g., cabinet maker).
Significant changes in the occupations of Amish men clearly occurred between 1973 and 1993 (Table 1). In particular, there were decreases in the percentages of farmers and shop workers and increases in the percentages of men working in a factory, the lumber industry and construction. Although the percentage of men who farm decreased by about 50% between 1973 and 1993, the actual number of farmers remained fairly constant, suggesting that the maximum capacity of the settlement to support farming families may have been attained by 1973. Finally, the total number of different occupational labels provided in each directory increased from 104 in the 1973 directory to 168 in the 1993 directory, or by about 62%, which suggests that the transition from agriculture to wage labor has been associated with an increased diversity of occupations.
Of the 1631 families listed in the directory for the Geauga settlement in 1993, 674 families had at least one adult child, for a total of 3768 adult children. Of these, 83.3% were reported to be Amish and living in the Geauga settlement; 8.5% were Amish but living in another settlement; and 8.3% were non-Amish. Expressed differently, 27% of all families with at least one adult child had at least one child migrate to another Amish settlement, while 22% had at least one non-Amish child. There are no significant differences in the proportion between males and females who are Amish and living in the Geauga settlement (males: 82.8%; females: 83.8%; (2 = 0.7, p > .05); Amish and living in another settlement (males: 8.7%; females: 8.2%; (2 = 0.3, p > .05); or non-Amish (males: 8.5%; females: 8.0%; (2 = 0.3, p > .05).
Relationship Between Leaving the Church and Out-migration
There are fundamental differences between the decision-making processes that lead some adult children to leave the Amish church and others to migrate to another Amish settlement. Nevertheless, those families with at least one non-Amish child are slightly more than twice as likely as those who have had no children leave the church to also have at least one child who has migrated to another settlement (Table 2). One possible explanation for this somewhat surprising finding is that both leaving the church and out-migration reflect dissatisfaction to some extent (although very different kinds of dissatisfaction) with the situation within the family, congregation and/or settlement. Some adults apparently deal with this dissatisfaction by migrating to another Amish community, while others decide to leave the church.
RELATIONSHIP BETWEEN LEAVING THE AMISH CHURCH AND MIGRATION TO ANOTHER AMISH SETTLEMENT
Number of Migrant Children in Family
0 > 1
Number of non-Amish ______________ ______________
children in family N % N %
0 412 78.3 114 21.7
> 1 79 53.4 69 46.6
(2 = 36.3, p < .001
Changes in Rates over Time
Table 3 breaks the sample down into four 10-year birth cohorts covering the period 1928 to 1967, and thus includes individuals who vary in age from 25 to 65 years of age. No older birth cohorts are included because they contain too few individuals for statistical analysis. No younger individuals are included because many of them have not yet decided whether they will join the church. The cohorts were thus selected so that most, if not all, of the youngest adult children described in Table 3 have decided whether they will join the church.
The rates at which adult children have made decisions about migrating to another Amish settlement and leaving the church have changed over time (Table 3). There was a substantial increase (by 40% in males and 32% in females) between 1928 and 1967 in the rate at which Amish adult children have decided to remain Amish and live in the Geauga settlement. Similar retention rates were reported for the Elkhart-LaGrange settlement for the period 1920-1969. The increase over time in the percentage of adult children remaining Amish and living in the Geauga settlement is partially a function of a slight decrease in out-migration rates over time, but is primarily the result of a substantial decrease in the rate at which adults have decided to leave the Amish church-a pattern similar to the one found in the Elkhart-LaGrange settlement.
STATUS OF ADULT CHILDREN
(a) Amish and living in the Geauga settlement (ALG); (b) Amish and living elsewhere (ALE); or (c) non-Amish (NA)
ALG ALE NA
Birth Cohort N % N % N %
1 January 1928 –
31 December 1937 44 47.8 18 19.6 30 32.6
1 January 1938 –
31 December 1947 167 67.1 41 16.5 41 16.5
1 January 1948 –
31 December 1957 289 72.8 56 14.1 52 13.1
1 January 1958 –
31 December 1967 460 87.3 41 7.8 26 4.9
ALG ALE NA
Birth Cohort N % N % N %
1 January 1928 –
31 December 1937 49 55.1 13 14.6 27 30.3
1 January 1938 –
31 December 1947 140 64.2 38 17.4 40 18.3
1 January 1948 –
31 December 1957 300 77.3 53 13.7 35 9.0
1 January 1958 –
31 December 1967 445 87.4 39 7.7 25 4.9
Considering first out-migration rates, the sharp decline in out-migration rates in adult children in the 1958-1967 cohort probably reflects the fact that this is a fairly young cohort of adults (25-35 years), so that not all of those who will eventually migrate have as yet done so. If one considers only the three older cohorts (some of whom may migrate in the future but most of whom will probably remain within the Geauga settlement), there was a small but statistically insignificant decrease between 1928 and 1957 in the rate of out-migration of both males ((2 = 1.9, p > .05) and females ((2 = 1.6, p > .05), with no difference between the sexes ((2 = 0.5, p > .05). In other words, the out-migration rate decreased slightly from 1928 to 1957 but essentially remained at a consistently high level of 14-20% of all adult children. Out-migration will thus clearly play an important role in the population dynamics of the Geauga settlement for many years.
The Amish have a long history of establishing new settlements. The existence of relatively high levels of out-migration from the Geauga settlement is therefore not surprising. The out-migration rates for the Geauga settlement are similar to those reported for the Lancaster settlement, but they are substantially higher than those reported for the Elkhart-LaGrange settlement, where the rate of out-migration decreased from about 8% to 3% between 1920 and 1969. Also, women were slightly more likely than men to migrate from the Elkhart-LaGrange settlement, which was not the case in the present study.
Although there are multiple push and pull factors associated with the relatively high rates of Amish migration to new settlements, most scholars have argued that the increasing cost of land in the major settlements, in association with a rapid increase in population size and the availability of less expensive farm land elsewhere, are the primary push and pull factors motivating Amish migration. The fact that the absolute number of farmers in the Geauga settlement remained fairly constant between 1973 and 1993 is consistent with this hypothesis, since it suggests that most of the available farm land is being utilized. However, given that the size of the Geauga settlement has been doubling in size every 20-30 years, this hypothesis would lead one to predict a corresponding increase in out-migration rates. In other words, as population size has increased and land within the settlement has become both less available and more expensive, this explanation would predict a corresponding increase in out-migration rates. However, there has actually been a slight decrease in out-migration rates over time (Table 3), suggesting that an inability to purchase affordable farm land within the Geauga settlement has not been the dominant factor in motivating families to migrate to a new settlement. The pattern of slightly decreasing out-migration rates in the Elkhart-LaGrange settlement is also consistent with this conclusion.
The reason that increasing population size has not been associated with an increase in out-migration rates is almost certainly because of the transition away from farming and towards various wage labor occupations (Table 1). One of the benefits of wage labor that has generally not been recognized is that wage laborers require much less land, and land of lower quality, than do farmers. While an Amish farmer generally requires 40-60 acres of reasonably fertile farmland, a wage laborer requires only 2-4 acres of grazing land for the family’s horses. As a result, the transition from farming to wage labor has permitted a much higher population density within Amish settlements. As will be discussed shortly, the value placed on farming appears to be changing in at least the Geauga settlement, so that for some families, farming was never really a serious option. Although farming is clearly a desirable occupation for some families, the fact that the out-migration rate actually decreased slightly despite a substantial increase in population size suggests that, faced with the choice of either staying in their home settlement as wage laborers or moving to a new settlement as farmers, many families choose to become wage laborers. Given that out-migration results in a substantial decrease in the social support that is such an integral part of Amish life, such a decision is not surprising. However, the option of working as a wage laborer, particularly as a factory worker, did not exist in the past, when such work would have been grounds for excommunication.
Many examples have been provided of the ability of the Amish to successfully adapt to various changes in world conditions while retaining their core beliefs. We argue that the change in the opinion of the Amish as to the appropriateness of wage labor, thus allowing families to stay in their home settlement, is another example of adaptive responses by the Amish.
Although out-migration rates are not increasing in response to increasing population density, they have nevertheless been maintained at a fairly high level. This then raises the question of what has been motivating 14-20% of all adult children to out-migrate every generation? We have no data on the reasons why some families choose to migrate from the Geauga settlement, but we assume that the same factors reported elsewhere are also operating in the Geauga settlement. Thus, some families certainly migrate in search of less expensive farm land. However, this may not be the only reason for migrating. We suggest that the second reason generally given for Amish out-migration-that is, the desire to live under either a more or, especially, a less conservative Ordnung than is available in the birth settlement-may also be playing an important role in the decision of some families to migrate from the Geauga settlement. Out-migration has long been recognized as one mechanism by which internal conflict revolving around the Ordnung has been reduced in Amish society.  Since farming is often associated with a more conservative Amish ideology, some proportion-and possibly a high proportion-of those who migrate because they wish to live under a more conservative Ordnung may also migrate to an area where they can farm. In such a case, however, one could argue that agriculture is seen as one component of living a more conservative lifestyle, or a returning to the roots of Amish religion, rather than as a primary determinant of the migration.
Although a slight decrease in the rate of out-migration is partially responsible for the significant increase over time in the proportion of adult children who have decided to remain within the Geauga settlement, the primary factor responsible for this trend is a substantial decrease over time in the rate at which young adults have chosen to leave the church-decreasing from about 30% in the oldest cohort to about 5% in the youngest cohort. The decrease in the rate of leaving the church over time was statistically significant in both males ((2 = 71.7, p < .001) and females ((2 = 69.0, p .05). More individuals belonging to these cohorts, particularly the youngest one, may leave the church in the future. However, as noted earlier, most individuals who leave the church do so as young adults. As a result, most of the individuals in even the 1958-1967 cohort who are going to leave the church have probably already done so.
The magnitude of the decrease over time in the rate at which individuals from the Geauga settlement have left the church is similar to that in the Elkhart-LaGrange settlement, where the rate of defection decreased from about 18% to 5% between 1920 and 1969. On the other hand, males were about twice as likely as females to leave the church in the Elkhart-LaGrange settlement,  while no difference between the sexes was found in the present study. Ericksen and colleagues found loss rates of 18-24% from the Lancaster settlement between 1900 and 1939.  In a more recent study, Huntington reported that interviews in 1962 and 1972 with 61 women in the Lancaster settlement indicated that fewer than 10% of their children had left the church. Nagata reported that 51% of adults left the church in a small Illinois settlement between 1936 and 1966.  However, this estimate was based on only 36 families, and the study was conducted in a small settlement that had recently undergone a great deal of factionalism. Sociologist Donald Kraybill reported a loss rate of 12% between 1953 and 1974 from the Holmes settlement. Troyer and Willoughby estimated that 23% of the children born in the Holmes settlement between 1935 and 1944 left the church, as compared to 15% of those born between 1945 and 1954. Thus, when comparing similar time periods, the loss rates found in the present study are generally similar to those reported elsewhere.
The decreasing rate at which Amish youth have left the Amish church is particularly surprising because it occurs at the same time as the transition from farming to wage labor, which has resulted in Amish youth having increasing exposure to the non-Amish world. This is particularly true for men, especially those working in factories. However, it is also true of women who, even though they have fewer occupational opportunities than males, frequently work outside the Amish community prior to marriage. Thus, both boys and girls are exposed to the non-Amish and their lifestyle at the same time in their lives when they must make the decision of whether or not they will join the church.
Given that Amish youth are increasingly less isolated from the non-Amish world, how then does one explain a decrease in defection rates over time? It is possible of course that Amish parents have become increasingly effective over time in transmitting Amish values to their children. However, although the Amish have changed over the last 50 years, the non-Amish world has changed to a much greater extent. As a result, the ways of life of the Amish and the non-Amish have become increasingly different. John A. Hostetler has convincingly argued that one factor involved in the continued survival of the Amish is their effective use of cultural symbols to distinguish themselves from others. If so, perhaps the increasing differentiation between Amish and non-Amish society has resulted in these symbols becoming increasingly effective over time in defining Amish identity. In addition, as a result of their greater exposure to the non-Amish world, Amish youth may now be better able to evaluate the positive and negative aspects of both an Amish and a non-Amish way of life. In at least the Geauga settlement most young adults clearly decide that the advantages of remaining Amish far outweigh the advantages of becoming non-Amish.
In summary, there has been a statistically significant increase over time in the rate at which adult Amish children in the Geauga settlement have decided to remain Amish and live in the Geauga settlement, primarily as a result of a significant decrease over time in the rate at which adult children have decided to leave the Amish faith. Although out-migration rates have decreased only slightly over time, they have nevertheless been maintained at a fairly high level. The implications of these patterns for the Amish will be discussed in a later section, after first considering the impact of various factors on the decisions to out-migrate or to leave the Amish faith.
A great deal of emphasis has been placed by both researchers and the Amish themselves on the importance of farming in the maintenance of the Amish way of life, with particular importance given to the extent to which small family-owned farms facilitate the maintenance of isolation from the non-Amish world and to the advantages of having both parents directly involved with all aspects of child rearing. One would therefore predict that farming families would be less likely to have children leave the church. On the other hand, children of farmers might be expected to be more likely to wish to farm themselves, making them more likely to migrate to an area where land is more affordable. The results of a comparison between farmers and non-farmers (shop, factory, lumber, construction, other) was consistent with the first hypothesis but not the second (Table 4a). In particular, about 15% of farming families had at least one non-Amish child, while the equivalent value for non-farming families was about 23%, a difference that is statistically significant (p .05), with about 25% of both groups having at least one child who had out-migrated.
An examination of Table 4b, which compares rates of out-migrating and leaving the church by five occupational categories, suggests that the comparison between farmers and non-farmers is somewhat more complex than it at first appears. In particular, the significant differences between occupational categories in both rates are strongly influenced by the “other” category. For example, families with a father’s occupation of “other” were about 2-3 times more likely than farmers to have at least one adult child who has migrated or who has left the church. The high rates of out-migration and leaving the church by the children of men whose occupation was classified as “other” could simply be due to sampling error. On the other hand, one consequence of the core Amish belief in humility is an emphasis on conformity. Some fathers who choose an atypical occupation are almost certainly doing so for purely economic reasons (i.e., the job was available). Perhaps others, however, are purposely choosing an atypical occupation, in which case they are exhibiting some level of nonconformity.
RELATIONSHIP BETWEEN OCCUPATIONAL CATEGORY AND WHETHER A FAMILY HAS HAD ANY ADULT CHILDREN MIGRATE TO ANOTHER AMISH SETTLEMENT OR LEAVE THE AMISH CHURCH
Number of migrant children Number of non-Amish children
in family in family
0 > 1 0 > 1
N % N % N % N %
a. Farming Status
Farming 104 75.4 34 24.6 117 84.8 21 15.2
Non-farming 382 72.6 144 27.4 403 76.6 123 23.4
(2 = 0.4, p > .05 (2 = 4.3, p < .05
b. Occupational Category
Farming 104 75.4 34 24.6 117 84.8 21 15.2
Shop 55 69.6 24 30.4 63 79.7 16 20.3
Factory 147 88.6 19 11.4 145 87.3 21 12.7
or Lumber Yard
Construction 86 78.9 23 21.1 96 88.1 13 11.9
Other 94 54.7 78 45.3 99 57.6 73 42.4
(2 = 52.8, p < .05 (2 = 61.2, p < .05
If so, it would not be surprising if such families also differ from other families in ways that might influence the likelihood of their offspring migrating or leaving the church. For example, perhaps a higher proportion of such families are less strict in remaining separate from the world, so that their children reach adulthood with a somewhat different world view than other Amish, making them more likely to reject the Amish way of life. In any case, the potential impact of the choice of an atypical occupation deserves investigation.
If the men in the “other” category are excluded, there are significant differences between the other four occupational categories in the proportion of families with offspring who have migrated ((2 = 14.6, p < .05). In particular, both farmers and shop workers are 1.5-2 times more likely to have at least one of their adult children migrate to a new settlement than are men who work away from home. That children raised in a farming environment are somewhat more likely to migrate to a new settlement is less surprising than the fact that adult children of shop owners have the second highest rate of migration to another settlement. Perhaps the skills that children learn in their father's shop make it easier for them than for other non-farmers to migrate to another settlement, since they possess skills that are both highly portable and essential (such as the manufacture of buggies and specialized farming equipment) to the long-term success of any Amish community. In other words, their decision to migrate might be facilitated by the belief that they will be able to provide for their family by filling a needed economic niche.
Removal of the “other” category also substantially changes our understanding of the influence of occupational category on leaving the church. When the “other” category is removed, there are no longer significant differences between occupational categories in the proportion of families with non-Amish offspring ((2 = 3.2, p > .05). In fact, the percentage of families who have had at least one child leave the church is actually lower in families in which the father works in a factory, lumber yard or construction than it is for farming families. Ericksen and colleagues, on the other hand, found that the children of non-farmers in the Lancaster settlement were 5.7 times more likely to leave the church than the children of farmers.  They argued that men who run their own farms have been more successful than others at saving the capital required to finance the purchase of land for their children, so that their children are more likely to remain Amish. Meyers, on the other hand, found slightly lower defection rates in farmers than non-farmers in the Elkhart-Lagrange settlement but the differences were statistically insignificant. Based on the similarity in defection rates between farmers and non-farmers, Meyers suggested that the Amish see wage labor as a means to an end (providing for their family) but not a source of identity, with the latter being provided by their religion and community. The results of the present study provide even stronger support for this hypothesis, since men who work away from home (factory, lumber, construction) are actually less likely than farmers to have their children leave the church. The importance of farming thus appears to differ substantially between the Lancaster settlement on the one hand and the Elkhart-LaGrange and Geauga settlements on the other. It is not clear if this difference is due to the different time periods during which these studies were conducted or if it reflects the fact that agriculture has long been less important in the Elkhart-LaGrange and Geauga settlements than in the Lancaster settlement.
Father a Religious Leader
Amish religious leaders (bishops, ministers, deacons) are chosen by lot, but only those men who are seen by the congregation as exemplifying Amish values are nominated for these positions. For this reason and since religious leaders play a central role in all aspects of Amish life and are expected to set an example for others, one might expect religious leaders, as a group, to come closer to meeting Amish ideals than others. This is in fact the case for the ideal of large family size, since religious leaders tend to have higher fertilities than non-religious leaders. However, there were insignificant differences between religious leaders and others in both the rate of out-migration ((2 = 1.5, p > .05) and leaving the church ((2 = 1.2, p > .05). In particular, 30.8% of the families of religious leaders had at least one adult child who migrated to another Amish settlement, while the equivalent value for the families of non-religious leaders was 25.9%. There was a similar non-significant difference between families with respect to leaving the church, with 18.9% of the families of religious leaders having at least one child leave the church, as compared to 23.0% of the families of non-religious leaders. Thus, in at least this respect, religious leaders truly are like everyone else.
INFLUENCE OF THE NUMBER OF ADULT CHILDREN IN A FAMILY ON WHETHER OR NOT A FAMILY HAS HAD AN ADULT CHILD MIGRATE TO ANOTHER AMISH SETTLEMENT OR LEAVE THE AMISH CHURCH
Number of migrant
children in family
0 > 1 0 > 1
Number of adult __________ ___________ __________ ___________
children in family N % N % N % N %
1 – 3 199 94.8 11 5.2
4 – 6 179 79.9 45 20.1
> 7 113 47.1 127 52.9
(2 = 137.2, p < .001
Number of non-Amish
children in family
0 > 1
Number of adult __________ ___________ ___________
children in family N % N %
1 – 3 195 92.9 15 7.1
4 – 6 186 83.0 38 17.0
> 7 145 60.4 95 39.6
(2 = 73.7, p < .001
Family Size and Parity
Given that neither leaving the church nor out-migration are particularly frequent events, it would not be surprising if families with more adult children were more likely to have at least one of them decide either to leave the church or to migrate to another settlement. In fact, there is a highly significant relationship between the number of adult children in the family and the likelihood that at least one child has migrated to another Amish settlement or left the Amish church (Table 5, p < .001). For example, the probability that a family with 7 or more adult children has had at least one child migrate to another settlement is about 10 times greater than that for a family with 1 to 3 adult children. Similarly, the likelihood that a family with 7 or more children has had at least one child leave the church also increases with the number of adult children, with large families having about 5.5 times greater chance than small families of having a child leave the church.
The likelihood of migrating or leaving the church might also be influenced by parity, or birth order. The percentage of individuals who migrate or leave the church relative to the total number of adult children of that parity is described in Table 6. For example, of all first-born adult children (parity = 1), 9.7% have migrated to another Amish settlement and 8.2% are non-Amish. After controlling for the total number of adult children at each parity, the percentage of adult children migrating or leaving the church is reasonably constant through the first 10-11 parities and then becomes somewhat erratic due to the small sample sizes. This perception of no differences between parities in these rates is supported by the results of a Kolmogorov-Smirnov test applied to the rates for the first 11 parities. This test indicates that the distribution of rates does not differ significantly from a uniform distribution for migration (z= 1.2, p > .05) or leaving the church (z=0.6, p > .05). In other words, low parity offspring are no more (or less) likely to migrate or leave the church than high parity offspring. These results differ slightly from those for the Elkhart-LaGrange settlement, where Meyers concluded that parity had a small effect on likelihood of leaving the church, with older siblings being somewhat more likely to leave the church than younger siblings. 
To what extent might one child’s decisions influence the decisions of another? In particular, is an adult child more likely to migrate or leave the church if an older adult sib has already done so? In fact, if the first-born child does not migrate, at least one later born child migrates in only 19.6% of families, while if the first-born child does migrate, at least one later born child migrates in 50.8% of families-a difference which is statistically significant ((2 = 31.7, p < .05). A similar but even stronger pattern exists with respect to leaving the church. If the first-born child does not leave the church, at least one later born child leaves the church in only 15.1% of families, while if the first-born child does leave the church at least one later born child leaves in 60.4% of families ((2 = 65.8, p < .05). In other words, if the first-born child migrates, later born sibs are 2.6 times more likely to migrate than if the first-born child does not migrate. Similarly, if the first-born child leaves the church, later born sibs are 4.0 times more likely to leave the church than if the first-born child does not leave the church.
EFFECT OF PARITY ON WHETHER OR NOT A FAMILY HAS HAD AN ADULT CHILD MIGRATE TO ANOTHER AMISH SETTLEMENT OR LEAVE THE AMISH CHURCH
Migrate Leave the Church
Total adult Total adult % of total Total
children of migrants of who non-Amish at % of total
Parity this parity this parity migrated this parity non-Amish
1 649 63 9.7 53 8.2
2 566 47 8.3 51 9.0
3 521 42 8.1 50 9.6
4 458 38 8.3 35 7.6
5 363 29 8.0 27 7.4
6 303 18 5.9 26 8.6
7 241 27 11.2 23 9.5
8 185 15 8.1 13 7.0
9 148 12 8.1 10 6.8
10 110 14 12.7 12 10.9
11 83 7 8.4 5 6.0
12 56 1 1.8 3 5.4
13 39 1 2.6 1 2.6
14 27 3 11.1 1 3.7
15 12 3 25.0 0 0.0
16 5 0 0.0 0 0.0
17 2 0 0.0 0 0.0
The importance of the behavior of older sibs in predicting the decisions of younger sibs is further emphasized in Tables 7 and 8, which examine the probability of a child migrating or leaving the church as a function of the behavior of the next oldest child. Through at least the seventh-born child in a family (after which sample sizes become too small for these analyses), if one sib migrates or leaves the church the very next born child is significantly more likely to do the same. For example, if the first-born child migrates, then the second child has migrated in 15.9% of families, but if the first-born child does not migrate, the second child has migrated in only 6.2% of families. Thus, when considering parities 1 through 7, if one child has migrated, there is a 2.6-3.5 times greater chance that the next born child will also migrate, and if one child has left the church there is a 4.0-19.3 times greater chance that the next born child will leave the church.
EFFECT OF ONE ADULT CHILD MIGRATING ON THE PROBABILITY OF THE NEXT ADULT CHILD MIGRATING
Previous Sib a Migrant Previous Sib Not a Migrant
Total Adult % of Total Adult % of
Parity Migrants Sibs Total Migrants Sibs Total (2
2 10 63 15.9 37 599 6.2 8.1*
3 9 47 19.1 33 598 5.5 13.3*
4 6 40 15.0 32 575 5.6 5.7*
5 8 36 22.2 21 521 4.0 22.6*
6 3 27 11.1 15 476 3.2 4.7*
7 3 15 20.0 24 424 5.7 5.2*
There are several possible explanations for these findings. First, with respect to migration, although there is no quantitative data, we think that sibs often migrate to a settlement where another sib has already migrated. Perhaps a younger sib is more likely to migrate to a new settlement-always a risky venture, given the high percentage of failed settlements-if another sib, who can be expected to provide assistance in locating a suitable homestead as well as provide social, economic and psychological support, is already resident. Similarly, the existence of an older sib who has left the church, particularly if the individual left prior to joining the church, might provide a younger sib with a clearer understanding of both the advantages and disadvantages of leaving the church.
EFFECT OF ONE ADULT CHILD LEAVING THE CHURCH ON THE PROBABILITY OF THE NEXT ADULT CHILD LEAVING THE CHURCH
Previous sib non-Amish Previous sib Amish
Total Adult % of Total Adult % of
Parity Migrants Sibs Total Migrants Sibs Total (2
2 13 53 24.5 38 621 6.1 25.5*
3 19 51 37.3 31 623 5.0 73.5*
4 18 50 36.0 17 624 2.7 98.1*
5 9 35 25.7 18 639 2.8 47.6*
6 9 27 33.3 17 647 2.6 51.0*
7 10 26 38.5 13 648 2.0 61.4*
* p < .05
In addition, if the older sib has successfully adapted to a non-Amish life, close observation of someone who has made this transition might influence someone who is wavering in their decision. An alternative explanation, and one that will be more fully examined in the next section, is that this pattern may reflect two or more sibs reacting in a similar fashion to the same family environment.
IS LEAVING THE CHURCH CONCENTRATED WITHIN FAMILIES’
Do some families have characteristics that contribute in some way to their children deciding to leave the church? If so, one might expect to find a concentration of this behavior within families. Expressed differently, are some Amish parents less able than most in transmitting key Amish values? The fact that only 148 out of 674 families with adult children, or 22%, have had at least one child leave the church, and the fact that adult children are more likely to leave the church if an older sib has already done so (Table 8) both suggest that there is some concentration within families of the tendency to leave the church.
The distribution of the 148 families who have had at least one child leave the church by the number of their children who have left the church is provided in Table 9. Of the families that have had at least one child leave the church, slightly more than 50% had only one child leave the church, accounting for 25% of all children who have left the church. Given that no other children in these families decided to leave the church, this finding is not consistent with the expectations of a concentration of this behavior within families. On the other hand, some families have been very unsuccessful in retaining their children, since about 37% of the children who have left the church came from the 15% of these families who had 4 or more children who left the church. The loss of 4 or more offspring by these unsuccessful families is not simply a matter of their having more offspring than other families. Families who have had 1 to 3 children leave the church lost an average of 25.3% of all offspring, as compared to 61.6% for those families who lost 4 or more children.
DISTRIBUTION OF THOSE FAMILIES WHO HAVE AT LEAST ONE NON-AMISH ADULT CHILD
Families with non-Amish Total number of
children non-Amish children
Number of non-Amish
children in family N % N %
1 78 52.7 78 25.1
2 27 18.2 54 17.4
3 21 14.2 63 20.3
4 8 5.4 32 10.3
5 6 4.1 30 9.6
6 6 4.1 36 11.6
7 1 0.7 7 2.3
8 0 0.0 0 0.0
9 0 0.0 0 0.0
10 0 0.0 0 0.0
11 1 0.7 11 3.5
It is likely that these relatively few families are primarily responsible for the increased risk of a sib leaving the church if the next oldest sib has left the church.
Some evidence for a clustering of children who leave the church within some families has also been found in the Lancaster settlement. Meyers, however, concluded that the evidence for the Elkhart-LaGrange settlement did not support this conclusion.  On the other hand, Meyers found that only 21% of all families had had at least one child leave the church, which, we have argued, suggests at least some clustering. It thus appears that there is some concentration within families of children who leave the church due primarily to the existence of a relatively small number of families that tend to have a high proportion of their children leave the church. Something about the child-rearing practices and general family environment of these families may be at least partially responsible for their children being more likely to decide to leave the church. Presumably these families are simply unable to transmit Amish values to their children as successfully as the vast majority of families.
SUMMARY AND IMPLICATIONS FOR THE FUTURE OF THE AMISH
We have described the patterns of out-migration and leaving the church in the Geauga settlement of northeast Ohio. Relatively high out-migration rates (14-20% of all adult children) have been maintained in the Geauga settlement. The rate at which adult children decide to leave the church, on the other hand, has decreased substantially over time, despite the fact that Amish youth have had increasing contact with the non-Amish world. One cause of the increasing contact with the non-Amish world has been an increase in the proportion of Amish adults who work as wage laborers, rather than as farmers. This occupational transition is not associated with an increasing rate at which children leave the church and, in fact, at least in the Geauga settlement appears to be associated with a decreasing rate of leaving the church. Finally, there does appear to be some concentration of leaving the church within some families, but most families successfully raise their children to be Amish.
Finally, although the ability of the Old Order Amish to maintain their cultural identity while living in the midst of U.S. culture is well documented, there has long been speculation about the effect of the transition from small-scale family farming to wage labor on the ability of the Amish to continue to maintain their identity in the future.  These concerns revolve primarily around the impact on core values of an increased contact with the non-Amish world; the need for mothers to become the primary child care provider in families in which the father works away from home; the increased leisure time for men working in wage labor and for their children who no longer have farm chores; the increased income and benefits associated with wage labor; and the potential that wage labor will lead to socioeconomic stratification. The fear is that these changes will result in fundamental changes in the nature of the Old Order Amish belief system and may, as a result, ultimately lead to their demise as a distinct subculture.
Such discussions have been going on for decades. And yet the Old Order Amish church is not only still here but remains very strong, though it is certainly true that their survival has required changes in Amish lifestyle and world view, such as the increasing acceptability of wage labor occupations. Although Meyers has expressed some concern about the long-term impact of these changes, he has also argued that, at least thus far, these changes have been accomplished without any substantial detrimental impact on core Amish values. The findings of the present study are consistent with this conclusion. In particular, since almost all Amish youth decide to remain Amish and since wage laborers do not have higher rates of defection than farmers, it appears that, at least for the present, these economic and related changes have not been socially destructive. On the other hand, there is some question as to whether this will continue to be true in the future, as communities lose the older members of their community who were raised within a farming environment and children are increasingly raised by parents who themselves never farmed. In our view this question may have already been answered in the Geauga settlement, which has a fairly long history of wage labor employment, so that the parents of many of the younger adults in the community were themselves wage laborers throughout adulthood. The fact that the lowest rates of defection are found in the cohort of youngest adults suggests that there has not been a weakening of core Amish values within the Geauga settlement. In other words, the mechanisms required to maintain the separateness necessary for the long-term survival of the Old Order Amish are still effective, despite increasing contact with non-Amish society.
. A third key decision made by almost all Amish adults is the choice of a spouse for life. Likely, this decision at least sometimes influences the decisions about joining the church and migrating to another settlement, but its importance cannot be evaluated with the available data.
Return to Text
. Hostetler, Amish Society, 369-71; David Luthy, “Amish Migration Patterns: 1972 -1992,” in The Amish Struggle With Modernity, eds. Donald B. Kraybill and Marc A. Olshan (Hanover: U. Press of New England, 1994), 243-59.
Return to Text
. Thomas J. Meyers, “Population Growth and Its Consequences in the Elkhart-LaGrange Old Order Amish Settlement,” MQR 65 (1991), 308-21; Thomas J. Meyers, “The Old Order Amish: To Remain in the Faith or To Leave,” MQR 68 (1994), 378-95.
Return to Text
. Louise S. Acheson, “Perinatal, Infant, and Child Death Rates Among the Old Order Amish,” American Journal of Epidemiology 139 (1994), 173-83; Harold E. Cross and Victor A. McKusick, “Amish Demography,” Social Demography 17 (1970), 83-101.
Return to Text
. Donald B. Kraybill and Steven M. Nolt, Amish Enterprise: From Plows to Profits (Baltimore: Johns Hopkins U. Press, 1995); George M. Kreps, Joseph F. Donnermeyer and Martha W. Kreps, “The Changing Occupational Structure of Amish Males,” Rural Sociology 59 (1994), 708-19; Thomas J. Meyers, “Lunch Pails and Factories,” in The Amish Struggle With Modernity, 165-81.
Return to Text
. The “shop” category may be underrepresented in both 1973 and 1993 as a result of men with shops providing an occupational label that was not clearly that of a shop worker and thus being placed in the “other” category.
Return to Text
. Julia A. Ericksen, Eugene P. Ericksen, John A. Hostetler, and Gertrude E. Huntington, “Fertility Patterns and Trends Among the Old Order Amish,” Rural Sociology 45 (1980), 49-68; Hostetler, Amish Society, 370-71; Luthy, “Amish Migration Patterns,” 243; Meyers, “Population Growth,” 313.
Return to Text
. J.A. Ericksen, E.P. Ericksen, and J.A. Hostetler, “The Cultivation of the Soil as a Moral Directive: Population Growth, Family Ties and the Maintenance of Community Among the Old Order Amish, Rural Sociology 45 (1980), 49-68; Kraybill, Riddle of Amish Culture, 196; Kraybill and Nolt, Amish Enterprise, 31; Luthy, “Amish Migration Patterns,” 243-44; Meyers, “Population Growth,” 312-13.
Return to Text
. Gertrude E. Huntington, “The Amish Family,” in Ethnic Families in America: Patterns and Variations, 3rd ed., ed. C. H. Mindel, R. W. Habenstein and R. Wright, Jr. (Englewood Cliffs: Prentice Hall, 1988), 367-99.
Return to Text
. Henry Troyer and Lee Willoughby, “Changing Occupational Patterns in the Holmes County Ohio Amish Community,” in Internal and External Perspectives in Amish and Mennonite Life, ed. Werner Enninger (Essen: Unipress, 1984), 52-80.
Return to Text
. There are many examples in literature produced by the Amish. One recent example is Gideon F. Stoltzfus, “Where are We Headed'” Family Life, Nov. 2001, 33-34, in which he tries to answer the question: “As the farmers become more and more a minority in many of our church districts, what effect is it having on the younger generation'”
Return to Text
. Although there is no quantitative data, wage laborers in the Geauga settlement generally seem to have greater incomes than farmers and are thus much more capable of accumulating capital to help their children get a start in their occupation of choice, including farming, than are farmers.
Return to Text
. Farming is still reported to be more highly valued than any other occupation in the Lancaster settlement, with many men working as wage laborers only long enough to accumulate sufficient capital to purchase a farm.-Huntington, “The Amish Family,” 380-81. In the Geauga settlement, however, no one occupation seems to be more highly valued than another; the key is whether or not the occupation allows a man to provide for his family. In addition, many men appear never to have had any desire to farm for a living.
Return to Text
. Completed fertilities in the Geauga settlement of 8.2 births for the families of religious leaders and 7.5 for non-religious leaders.-Greksa, L.P., unpublished data. Similar differences have been reported for the Holmes settlement.-Samson W. Wasao and Joseph F. Donnermeyer, “An Analysis of Factors Related to Parity among the Amish in Northeast Ohio,” Population Studies 50 (1996), 235-46.
Return to Text
. Lyle R. Fletcher, “The Amish People of Holmes County, Ohio: A Study in Human Geography” (M.A. Thesis, Ohio State U., 1932); Donald B. Kraybill, “Introduction,” in The Amish Struggle With Modernity, ed. Donald B. Kraybill and Marc A. Olshan (Hanover: U. Press of New England, 1994), 1-17; Donald B. Kraybill and Steven M. Nolt, “The Rise of Microenterprises,” The Amish Struggle With Modernity, 149-63; Martineau and MacQueen, “Occupational Differentiation Among the Old Order Amish,” 395-96; Marc A. Olshan, “The Opening of Amish Society: Cottage Industry as a Trojan Horse,” Human Organization 50 (1991), 378-84; Kreps et al., “The Changing Occupational Structure of Amish Males,” 716-18; Conrad L. Kanagy and Donald B. Kraybill, “The Rise of Entrepreneurship in Two Old Order Amish Communities,” MQR 70 (1996), 263-79; D. Yutzy, “The Decline of Orthodoxy Among the Amish,” Sociological Focus 2 (1968), 19-26.
Return to Text
. Joseph F. Donnermeyer, “Amish Society: An Overview,” The Journal of Multicultural Nursing and Health 3 (1997), 6-12.
The Mennonite Quarterly Review
Key Decisions in the Lives of the Old Order Amish
*Lawrence P. Greksa and Jill E. Korbin are both professors of anthropology at Case Western Reserve University.