The Interdependence of Vital Events:
Twentieth-Century Indian Kerala
S. I. Rajan and K. S. James
Tables
Studies that use short-run analyses of demographic variables to understand historical populations have gained predominance during the last two decades, particularly in the context of developed countries. Short-term analysis is often thought to provide greater insights into a population's demographic and economic history. In addition, short-run analysis is also considered to be a good measure to examine even a contemporary population. 1
In the short run, several important variables influencing vital events, such as age structure and social/cultural indicators, remain constant, thus limiting the number of exogenous and endogenous variables. Of the exogenous factors, economic variables ought to play an important role. Another important exogenous variable in the short run is weather. However, since adverse weather conditions [End Page 21] also directly affect living standards, they can be captured, to some extent, by economic variables.
Studies based on short-term analysis have mainly concentrated on how exogenous factors create fluctuations in vital events; the effect of endogenous variables on birth, death, and marriage seldom receive much attention. Since fertility, mortality, and nuptiality are sufficiently interconnected to cause fluctuations in one another, an analysis of their links can provide important information about the demographic behavior of specific population groups. 2
Analyses of short-term variations have clearly revealed that intra-yearly variations (seasonal) are much more intense than annual ones, conveying significant information about the social, economic, and physical environment of a population. Hence, this article analyzes short-term fluctuations in the vital events of two Catholic groups in Kerala--Syrian Catholics and Latin Catholics. Its objectives are twofold: to ascertain the extent of the seasonal (monthly) variation in the vital events of these two groups and to examine how the interactions between fertility, mortality, and nuptiality affect each group.3
Although the Syrian and Latin Catholics share a state and a faith, they are distinct in their historical evolution. Furthermore, since conclusive evidence suggests that the Syrian Catholics are more advanced in socioeconomic standing than the Latin Catholics, we deem it safe to propose that the fluctuation in vital events is more intense among Latin Catholics than among Syrian Catholics.
An Overview of Kerala
Kerala state is a narrow stretch of land bounded by the Arabian Sea in the west and the Western Ghats mountains in the east. Kerala state was formed in 1956 from a majority of the Malayalam-speaking areas of two princely states, Travancore and Cochin, and the Malabar area of Madras presidency. Hence, Keralites are generally known as "Malayalees." The state is divided into three broad natural physiographical divisions: (1) the lowland, which is mainly the coastal strip (elevation below 25 m); (2) the midland (elevation of 25 to 250 m); and (3) the highland (elevation greater than 250 m). The highland area features [End Page 22] mainly plantation crops, such as tea, coffee, and cardamom, and is sparsely populated, with a population density of 172 per sq. km. The midland supports a combination of commercial and food crops; it has a population density of 778 per sq. km. The lowland area, which grows mainly food crops (paddy rice), has the highest density, 1,385 per sq. km.
Kerala's impressive improvement in demographic variables has been discussed at length in development literature. Its fertility decline started by the early 1960s and reached replacement level by the late 1980s, though without any significant improvement in its material development. Mortality transition in the state is traced back even further to the beginning of this century. 4
The Origin of Christianity in Kerala
At present, Christians constitute nearly 20 percent of the total population in Kerala, compared with 2.6 percent in India as a whole. Christianity in Kerala is said to date from 52 A.D. when St. Thomas, according to legend, reached the Southwest coast of India and converted many into the new religion, even many upper-caste Hindus, or Brahmins. The name "Syrian" came to be applied to these early converts because they used the Syriac language for worship. 5
Notwithstanding the legend of these "St. Thomas Christians," Kerala had trade links with the Persian Gulf and Greeks even before Christian era, and evidence of an early connection between Malabar and the Persian church provides some confirmation that a church existed in Kerala. That a group of foreign settlers under the leadership of Thomas of Cane migrated to Kerala is indisputable, though whether it occurred in the middle of the fourth [End Page 23] century or much later is undecided. A particular sect of Syrian Christians who claim to be direct lineal descendents of Thomas of Cane exists in the state, but there is no record of Christian life in Kerala until the eighth century. The relationship with the church in Persia seems to have continued in one form or another.
At any rate, the tranquil life of Syrian Christians in Kerala was later to be disrupted when Vasco da Gama discovered a route from Europe to India via the Cape of Good Hope in 1498. The initial aim of the Portuguese was commercial, and the powerful, and prosperous, Syrian Christian community, which controlled much of the pepper trade, benefited from it. But by 1510, the Portuguese were trying to bring the Syrian Christians under the Roman Church and wrest them away from the Patriarch of East, who had been sending bishops to monitor the spiritual needs of this community. Friars began to penetrate into the local churches to conduct masses according to the Latin rite. Alexis de Menezes, the newly appointed archbishop of Goa, visited Kerala at the close of the sixteenth century to appoint a Jesuit as bishop and consolidate the Roman authority. The sole motive of the synod of Udayamperur in 1599 was the establishment of Roman doctrines. The Syrian Christians dared not oppose this interference for fear of losing their lucrative trading arrangement with the Portuguese. Hence, a majority of Syrians changed allegiance outright. Known as Syrian Catholics, they they currently constitute more than 40 percent of the Christian population.
Those Syrians who do not accept Roman supremacy--the second largest group of Christians--adhere to the Jacobite, or Orthodox, church, under the Patriarch of Antioch who resides in Damascus. A split in the Jacobite section during the nineteenth-century led to the formation of the Mar Thoma Syrian church, the third largest Christian group in the state. Other small groups have joined the Anglican church, Church of South India, Chaldean Christians, Anglo-Indians, etc.
The Economic and Social Life of Syrian and Latin Catholics in Kerala
Because Kerala is remote from the rest of India to some extent, its caste divisions are considerably different from those in other parts of the country. For one thing, it has no formal business caste group; the Syrian Christians have largely filled this role. However, they have also made deep inroads into the agricultural [End Page 24] sector, aided by the spurt in land transactions that followed the break-up of the joint-family system among the privileged Nair caste and the diversion of surplus money from trade to land purchase. According to the 1875 census, 23 percent of the total Christian male population were cultivators and 13 percent were laborers. Table 1 presents the distribution of cultivators and agricultural laborers by caste and community according to the 1931 census. 6
A cursory look at Table 1 reveals that nearly 15.6 percent of total earners among Syrian Christians are engaged as agricultural laborers. The percentage of cultivators in this community is higher, indicating that the Syrian Christians are not an economically homogeneous group. Nearly half of wet-land owners and 31.6 percent of dry-land owners possess less than 1 acre. The percentage of those who own more than 10 acres of wet land and dry land is 4.1, and 6.8, respectively. Clearly, Syrian Christian farmers do not possess much land, despite their elevated social rank. Although their social status is roughly similar to that of the Nairs, their economic status differs, since Nairs by and large do not work as agricultural laborers. 7
As upstanding members of society, the Syrian Christians tend to honor Hindu traditions. Not only do they participate in many Hindu festivals and give offering to the temples; they are even reluctant to baptize and accept low-caste Hindus into their religion. Hence, the growth of their community is conditioned mainly by the natural process of births and deaths. 8
The Latin Catholics are the descendants of coastal fishermen from the lower castes--the pulayas and parayas, who once were slaves. Although slavery per se was abolished in the mid-nineteenth century, serious discrimination, sanctioned by Hinduism, continued. Conversion to Latin Christianity played an important role in emancipating the less fortunate from the worst indignities of the caste system. Yet, unlike the privileged Syrian Catholics, the Latin Catholics retained no loyalty to the practices of [End Page 25] the Hindu religion. The influence of Western missionaries almost completely succeeded in coloring their outlook. 9
Nonetheless, the economic emancipation of this group was slow. A survey conducted in the Trivandrum archdiocese as late as 1992 revealed that the majority of Latin Catholics were laborers or fishermen. The same survey also indicated that in most of the Latin Catholic parishes, more than 50 percent of the houses were made of temporary materials like palm leaves and grass. 10
Syrian Catholics and Latin Catholics are not only distinct socially and economically; they also diverge religiously. They each have their own official hierarchies and practices. The upshot is that their differing characteristics and circumstances translate into different demographic profiles of birth, death, and marriage.
Sources of Data
The data for births, deaths, and marriages come from the parish registers of five Syrian Catholic parishes in and around Palai, Kottayam, and six Latin Catholic parishes in and around Trivandrum. The baptisms, burials, and marriages that the parish registers record for religious purposes provide data that are meant to correspond with births, deaths, and marriages. However, underestimation could cause serious problems in the short-run analysis of these vital events. Although all Catholic children in Kerala are supposed to receive their baptisms in church, parish [End Page 26] priests can conceivably fail to record them on occasion. Parents, too, can fail to make the compulsory report of a baptism done in another parish. We have encountered several such instances in our research. In an earlier era, the birth-baptism gap in Kerala was narrow (less than 7 days); most baptisms were taking place in the parishes of residence. Large-scale in- and out-migration have subsequently created discrepancies in the number of vital events. From his experience with British parish registers, Wrigley argued that the risk of a birth-baptism shortfall is greater than that of a burial-death or marriage shortfall, although the death of an infant before baptism might not be recorded in the register. 11
Analysis of the parish records in Kerala for both Syrian Catholics and Latin Catholics confirms their usefulness, but it also shows that the quality is better regarding the Syrian Catholics than the Latin Catholics, probably because the Latin Catholic data is complicated by the conversion of new people into the religion.
Method of Analysis
The data available from the parishes of Syrian and Latin Catholics gives only the total number of births, deaths, and marriages in broad outlines. However, because no other population figures are available for each period, we assume that the population remained constant in the short period.
To analyze seasonality, we calculated a monthly index number that corresponds to the total number of births in a period first by dividing the total number of events by 12, which gives the expected number of events in each month, assuming an even split of events. The index is the ratio of the actual events to the evenly distributed events for a particular month, multiplied by 100. 12
For the analysis of annual fluctuation, it is essential to eliminate long-term trends from the available data series. First we divided each series by a nine-year-centered moving average. This method effectively removes fluctuations lasting more than fifteen to twenty years that are associated with changes in the size and [End Page 27] age structure of the population. Another major advantage of such a procedure is that since the mean of the resulting series is nearly unity, its coefficient in the statistical analysis is equivalent to elasticity. The coefficient, therefore, conveys the precise magnitude of fluctuation in one variable due to the variation in the other. 13
A distributed lag model is employed to capture lagged influences affecting the interrelationship of births, deaths, and marriages. Accordingly, the dependent variable is regressed with the values of the independent variable for the current year, as well as for a few years earlier. This procedure identifies the impact of explanatory variables of the past on the contemporary dependent variable.
We employ a time lag of four years to signify the influence of an independent variable on the dependent variable for the next four years. The selection of a four-year lag is altogether arbitrary, not founded on any theoretical consideration. We also assume that the relationship between the dependent and independent variable is linear. The ordinary least-square method is utilized to estimate the coefficient of the explanatory variable on the dependent variable during the four lag years. 14
Seasonal Variation in Vital Events
In several industrialized countries, seasonal fluctuations in vital events are much more pronounced than annual variations. Seasonal variations in vital events, particularly births, are well acknowledged in all populations. It is still not clear why births in one month exceeded births in other months during the postwar years in the United States. The number of births in September, for example, exceeded the number of births in April by nearly 10 to 20 percent. In England, during the two centuries prior to 1750, the first four months of the year saw 10 to 20 percent more baptisms than would have been expected on the basis of an even split between months. In some of the parishes of Mumbai (Bombay), India, the percentage [End Page 28] of children born in the months of September to December was the highest and the percentage born in February was the lowest. 15
Seasonal variation occurs not only in births but also in burials and marriages. In Bombay, frequent burials were recorded during the monsoon season (36 percent), followed by winter (33 percent) and summer (31 percent), and more than 50 percent of marriages there take place in January, February, May, and November. In England, seasonality in marriage was more pronounced between the sixteenth and nineteenth centuries. However, the overall distribution peaked in early summer and autumn, followed by a trough in late summer and March. 16
Though there is consensus regarding the existence of fluctuation in vital events, there is no unanimity in the conclusions drawn from different studies with regard to its determinants. Climate is often cited as a crucial variable. Economic variables as well as customs and practices--including religious celebrations, festivals, and superstitions--seem to contribute to such fluctuations. Seasonal variation in marriage has been considered a signal for seasonal fluctuation in births. In the study of parishes in Bombay, marriages are found to be closely associated with the harvesting season. 17
Seasonality and Birth
The monthly index numbers corresponding to the total numbers of births in different time periods for Syrian and Latin Catholics are given in tables 2 and 3, respectively. An index value of 100 signifies the absence of any seasonal fluctuation; a value greater than 100 indicates a higher number of events and lower than 100 a smaller number.
Among Syrian Catholics, January has consistently accounted for more births every year than an evenly distribution in every month would yield. February and March also have a higher [End Page 29] number of births, except in two or three decades during the early part of the century. For most decades, April, May, June, September, November, and December have index values less than 100. July, August, and October always fluctuate around the index value of 100. December has undergone the most change, showing fewer births during the earlier decades compared with those from 1941 through 1980. [End Page 30]
The Latin Catholics show a considerably different pattern. For most decades, the first eight months of the year account for more births than the last four months. December records the lowest number of births; March records the highest index value during the last few decades. Of late, however, the seasonality in vital events seems to be receding, the index value not deviating considerably from 100, and seasonal fluctuations in birth have a lesser magnitude among this population than among Syrian Catholics.
Seasonality in birth depends primarily upon seasonality in conception. Assuming a gestation lag of nine months in every birth, the profusion of Syrian Catholic births in January, February, and March entail the previous April, May, and June as the months of conception. The scarcity of Syrian Catholic births during April and May means fewer conceptions in July and August. April, May, and early June are comparatively wealthy months in Kerala. Since these months precede the onset of the monsoon season, they provide considerable employment opportunities in the agricultural sector. Moreover, more marriages can take place in this period because April and May follow the forty-one days of Lent, according to the Catholic calendar, when marriages are generally avoided.
Conceptions in June, July, and August are rare among Syrian Catholics. However, Latin Catholics show a higher number of conceptions during these months. June, July, and August are the "poverty months" in Kerala; economic activities come to a standstill during this period due to heavy rain. Yet, these months seem to be favorable for conception among Latin Catholics, probably because this community earns its livelihood from fishing and other urban employment activities, which are not as seasonal as agricultural operations. Among Latin Catholics, the lowest number of conceptions occurs in January, February, and March when the monsoon season ends and fish are far less available than in other seasons.
That seasonality in births is more concentrated among Syrian Catholics than Latin Catholics, particularly in recent years, may seem strange in view of the fact that the standard of living of the Syrian Catholics is far higher than that of the Latin Catholics. But the higher standard of living of Syrian Catholics is not able to offset their seasonal occupation so far as the fluctuation in birth [End Page 31] is concerned. Contrary to several studies, standard of living is not the sole influence on birth patterns. Other factors contribute to the variations.
Seasonality and Marriage
Marriage is usually much more sensitive to the change in seasons than birth and death. Tables 4 and 5 present the index numbers of marriage for Syrian and Latin Catholics, respectively. 18
A substantial number of Syrian Catholic marriages take place in January and February, as well as in April and May; the drop-off in March signifies the observance of Lent. The number of marriages during the second half of the year is comparatively low. The index numbers from June to December usually show an index value of less than 100.
The monsoon season in Kerala, which starts in June and continues for three to four months, depresses the economic condition of the state and all but prevents marriage ceremonies. September and October have also registered few marriages, particularly in the first half of the twentieth century. The latter part of September and the first part of October fall in the month of Kanni, based on the Malayalam calendar, which is considered inauspicious for marriage. Since the Syrian Catholics follow many of the traditions, customs, and practices of the local Hindus, their reluctance to marry during these months is no surprise. However, education and modernization appear to have changed their attitude toward these customs. By the end of the century, the number of marriages in November had begun to increase, though not those in December, which, according to the Catholic calendar, falls within the twenty-five days of ritual before Christmas, when marriage celebrations are discouraged by Syrian Catholic authorities.
Unlike the Syrian Catholics, the Latin Catholics show their highest number of marriages during monsoon months. Since this group depends mainly on fishing, the monsoon months are their most prosperous months. Seasonality in marriages is also less in evidence in the case of Latin Catholics, even in the twenty-five days of Lent prior to Christmas, which seem to be less important to them than to Syrian Catholics. The Latin Catholics in Bombay, however, evince a pattern similar to that of the Syrian Catholics, [End Page 32] celebrating more than 50 percent of their marriages during January, February, and May. 19
Seasonality in marriage among those residing in the same area and maintaining the same living standards can differ according to their customs and practices. Hence, economic variables or weather may not necessarily be the sole explanation for seasonality in marriage. [End Page 33]
Seasonality and Death
The seasonality index of death for Syrian and Latin Catholics is similar to the indexes of birth and marriage (see tables 6 and 7). The data are available from 1911 onward for Latin Catholics and from 1931 onward for Syrian Catholics.
It is clear from Table 6 that the highest number of deaths among the Syrian Catholics occurs during the monsoon months (June through September) for all the decades without exception. January and September have an index value of c. 100, indicating what would be an average if all deaths were split evenly across twelve months throughout the decades. All other months have fewer deaths, the lowest being the month of November. The higher number of deaths during the monsoon months are expected since these months are known as the poverty months of the state due to torrential rains and infrequent job opportunities in agriculture. In addition to the hazard of curtailed job opportunities, these months also account for the spread of infectious disease, leading to a higher incidence of death.
Seasonality in deaths is only slight among Latin Catholics. The months of January through March bring the most deaths. Moreover, seasonality in the births and deaths of Latin Catholics seem to be linked, both numbers being high. The monsoon months cause them more deaths as well, although not to so great a degree as they do the Syrian Catholics. Studies carried out in the Bombay parishes, however, confirm that Latin Catholics there suffer the highest number of deaths in the monsoon season as well. The combination of so many deaths and births in those months may be an indication that neonatal deaths are more prevalent among Latin Catholics than among Syrian Catholics. 20
The analysis of seasonality demonstrates that the fluctuations in all the three vital events are more intense for Syrian Catholics than for Latin Catholics, despite the superior living standards of the former. In no way is this point intended to convey that the Latin Catholics were less poor in the past. On the contrary, it stresses the importance of such other variables as occupation and custom. Other studies have not given much weight to these variables concerning the fluctuation in vital events. [End Page 34]
The Interdependence of the Vital Events
Scholars have often observed that certain demographic events are linked in such a way as to cause fluctuations among themselves--for example, death and birth and death and marriage, as well as marriage and birth in the short run. The relationships between death and birth and marriage and birth are amply discussed in the fertility literature. The nature of these relationships provides a great deal of [End Page 35] information about the demographic and economic history of populations. 21
The Impact of Death on Birth
The impact of death on fertility has been a subject of debate, particularly in the historical demographic literature, for some time. According to Richards, death affects birth physiologically and behaviorally. In the former case, the death of a child interrupts lactation and increases the mother's risk of pregnancy; in the latter, parents often replace the child that died to achieve the desired family size. Hence, deaths are likely to affect fertility positively, at least with a lag effect. Yet, when the number of deaths is high, morbidity must be even higher, and disease can affect fecundity in a number of ways, ranging from foetal loss to reduction in coital frequency. The result is a negative association between death and birth, with a lag effect. Table 8 gives the lag coefficient of death on birth for Syrian and Latin Catholics. 22
Death and birth are associated negatively in the Syrian Catholic population. However, there is no significant relationship between these two variables in any lag period. The total variability in births explained by the independent variables is 7 percent. This low R 2 is partly due to the procedure employed to remove the long-term trend from the data. The lag sum is negative. The negative relationship between these two variables establishes that the desire to replace children after infant death is not predominant among Syrian Catholics. The data from the parishes of England and France also have shown a strong negative relationship between death and birth barring lag 2. 23
Nearly opposite results characterize the Latin Catholics; all the years show a positive relationship between death and birth. The lag sum is positive and much higher than that noted for Syrian Catholics. It indicates that Latin Catholics tended to have children to replace those who died. Since Latin Catholics tend to suffer more child mortality than Syrian Catholics because of their lower standard of living, their tendency toward child replacement to reach a desired family size would be more pronounced. Such replacements mainly take place with a lag of two or three years, as shown by the significant positive relationship observed at lag 3. [End Page 36]
The Effect of Death on Marriage
The association between death and marriage has a twofold outcome. An augmented mortality during crisis would create more widows and widowers, many of whom would remarry in the coming years, thereby boosting the number of marriages, sometimes with a time lag. In a country like France, where inheritance and remarriage are important paths to the material basis required for household formation, high mortality followed by high nuptiality is much the rule. Yet, Lee argues that since the number of marriages is a function of the marriageable population's size, increased mortality also leads toward a decrease in marriages. Lee also argues that marriages are often deferred in periods of crisis because of the death of a close relative. This postponement, however, has little long-run effect on the number of marriages. 24
It is noteworthy in this context that Kerala's considerable increase in deaths during the last century did not lead to a signi- ficant alteration in the age structure. Hence, it is reasonable to pro- pose a positive association of deaths and marriages, reflecting the theory that crises can cause marriages to increase. Table 9 shows the net effect of death on nuptiality for Syrian and Latin Catholics.
The relationship between death and nuptiality is generally positive for Syrian Catholics and Latin Catholics. The lag sum is positive and nearly the same for both the groups. The fact that the Latin Catholic population shows a significant positive relationship between death and marriage at lag 0 means that an [End Page 37] increase in deaths during crisis periods leads to an immediate increase in marriages, probably reflecting the incidence of remarriage in the population. Information about the extent of remarriage among Latin Catholics in Kerala is not available. However, the data from the Latin Catholic parishes in Bombay show a marked tendency toward remarriage in this community. The percentage of second marriages among Latin Catholics in Bombay was as high as 33 percent from 1869 to 1874; it declined to 15 percent from 1925 to 1935, 4 percent from 1950 to 1964, and finally to 2 percent from 1975 to 1984. 25
At lag 4, the relationship is positive and significant for Syrian Catholics; the relationship is negative but not significant for Latin Catholics. A marriage surge after a lag of four years among Syrian Catholics may be attributable to the specificity of the marriage ceremony in this part of the region. By tradition, parents arrange all Syrian Catholic marriages. Since the bride's parents incur the enormous expense of the celebration, including the dowry--they would be hard pressed to arrange another marriage for their daughter soon after a mortality crisis. It would take at least two to three years for them to recover from it and save enough for another ceremony. Studies undertaken in France and England show that death interrupts marriage at lag 0; a significant positive relationship obtains at lags 1 and 2. 26 [End Page 38]
Because Latin Catholics probably suffer a higher rate of mortality crises than Syrian Catholics do, it stands to reason that they would have more remarriages without any lag. By the same token, Syrian Catholics are in a better position to afford the luxury of postponing a marriage during crisis or after death.
The Impact of Nuptiality on Birth
A positive association of nuptiality and fertility is evident for most of the population. The potential for marriage to create overpopulation was voiced early by Malthus, who advocated celibacy and a moratorium on marriage to control birth rates. However, even though marriage and birth may not have a positive association in the short run where contraceptives are available, and couples practice family-planning methods before the first birth, Kerala's introduction to contraceptives in 1960 does not seem to have made much difference to the first-birth interval. Furthermore, because no evidence suggests that a significant number of illegitimate births occur within Kerala's Catholic community, a positive association between nuptiality and fertility seems likely in this population. Table 10 presents the coefficient of nuptiality on births for Syrian and Latin Catholics. 27
A notable influence of marriage on birth is significant in the same year, at lag 0, for both the groups, as well as at lag 1 for Latin Catholics. The relationship between the two variables after a lag of two years and more is not significant. In a society where the usage of contraceptives to plan birth is absent, conception normally follows marriage without delay. Hence, the positive correlation of marriage and birth in Kerala is hardly anomalous. Lee finds a similar association between marriage and birth elsewhere. 28
The main objectives of this study were to bring out the importance of endogenous forces in shaping fluctuation of vital events and thereby to understand the demographic and economic histories of the Syrian Catholics and the Latin Catholics in Kerala. Earlier studies of seasonality and annual fluctuations clearly exposed the causal importance of economic variables. Based on this precedent, we originally hypothesized that seasonal fluctuations in vital events would be more pronounced among the Latin Catholics [End Page 39] than among the Syrian Catholics. Nonetheless, contrary to the short-run analysis in several countries, we find just the opposite. The reasons for this anomaly primarily concern the differences between these two Catholic communities in socioeconomic, cultural, and religious practice and, more important, in occupation. The agricultural pursuits of Syrian Catholics are more seasonal than the fishing and other casual work that Latin Catholics do. Furthermore, the Syrian Catholic seasonality in marriage, which leads to fluctuations in birth, occurs mainly because of Hindu customs inherited from the larger society that the Latin Catholics do not follow. Hence, economic factors alone do not determine the fluctuations in vital events. Even at a relatively high standard of living, such as that of the Syrian Catholics, occupational characteristics and cultural traditions can have an immense effect on birth, marriage, and death.
The interconnections between vital events are, however, more pronounced among the Latin Catholics than the Syrian Catholics. Deaths have a significant positive effect on births, and the time lag indicates that the Latin Catholics are intent on replacing dead children, perhaps in response to a high incidence of infant mortality among them. In the Syrian Catholic population, the lag sum shows a negative effect.
Deaths also have a positive impact on marriages at lag 0 among Latin Catholics, reflecting the prevalence of remarriage. Among the Syrian Catholics, this relationship is positive but significant only at lag 4. Evidently, the Syrian Catholics postpone [End Page 40] marriage after a crisis or the death of a family member. Marriage has a significant positive impact on birth without any lag among both these groups; neither employs contraception to delay first birth.
This analysis of seasonality and the interdependence of vital events provides important information about these two Catholic groups. Unlike previous studies, this analysis reveals that living standards are not the sole yardstick for measuring short-term fluctuations in birth, marriage, and death. Sociocultural and religious practices must also enter into the investigation.
S. Irudaya Rajan is Associate Fellow, Centre for Developmental Studies, Trivandrum, India. K. S. James is Associate Fellow, Centre for Economic and Social Studies, Hyderabad, India. They are the authors of "Short Run Variation in Vital Events in Kerala: The Role of Economic Factors," Indian Economic and Social History Review, XXV (1998), 53-68.
Notes
1. A number of studies about both annual and seasonal fluctuations in the vital events of European populations are available. See, for instance, Richard J. Udry and Naomi M. Morris, "Seasonality of Coitus and Seasonality of Birth," Demography, IV (1967), 673-679; Ronald D. Lee, "Methods and Models for Analysing Historical Series of Births Deaths and Marriages," in idem (ed.), Population Patterns in the Past (New York, 1977); idem, "Models of Pre-industrial Population Dynamics, with Applications to England," in Charles Tilly (ed.), Historical Studies in Changing Fertility (Princeton, 1978); E. Anthony Wrigley and Roger S. Schofield, The Population History of England, 1541-1871: A Reconstruction (Cambridge, 1981); Tommy Bengtsson, G. Fridlizius, and R. Ohlsson (eds.), Pre-Industrial Population Change: The Mortality Decline and Short-Term Population Movements (Stockholm, 1984); Toni Richards, "Weather, Nutrition, and the Economy: Short-Run Fluctuations in Births, Deaths and Marriages, France 1740-1909," Demography, XX (1983), 197-212; Daniel A. Seiver, "Trends and Variation in the Seasonality of U.S. Fertility, 1947-1976," Demography, XII (1985), 89-100; Patrick R. Galloway, "Basic Patterns and Annual Variations in Fertility, Nuptiality, Mortality and Prices in Pre-industrial Europe," Population Studies, XLII (1988), 275-303; idem, Secular Changes in the Short-term Preventive, Positive and Temperature Checks to Population growth in Europe 1460-1909 (Berkeley, 1992); Kuriath Sebastian James and Sebastian Irudaya Rajan, "Short Run Variation in Vital Events in Kerala: The Role of Economic Factors," Indian Economic and Social History Review, XXV (1998), 53-68; Lee, "The Demographic Response to Economic Crisis in Historical and Contemporary Populations," Population Bulletin of United Nations, XXIX (1990), 1-15. The term, "short term," includes both annual and seasonal fluctuations.
2 Richards, "Weather," 201, tries to establish how endogenous variables in France caused fluctuations in vital events.
3 Wrigley and Schofield, Population History, 33-65.
4 Thippalassery N. Krishnan, "Demographic Transition in Kerala: Facts and Factors," Economic and Political Weekly, XI (1976), 1203-1224; John Ratcliffe, "Social Justice and Demographic Transition: Lessons from India's Kerala State," International Journal of Health Services, VIII (1978), 123-144; P. N. Mari Bhat and Rajan, "Demographic Transition in Kerala Revisited," Economic and Political Weekly, XXI (1990), 1957-1980; Kunniparampil Curien Zachariah, The Anomaly of Fertility Decline in India's Kerala State: A Field Investigation (Washington, D.C., 1984); Zachariah et al., Demographic Transition in Kerala in the 1980s (Thiruvanathapuram [Trivandrum], 1994); Zachariah and Rajan (eds.), Kerala's Demographic Transition: Determinants and Consequences (New Delhi, 1997); Rajan, Bhat, and Tim Dyson, "Onset of Mortality and Fertility Transition: New Evidence from Parish Records," in ibid., 112-120; P. G. Kesava Panikar and C. R. Soman, Health Status of Kerala: Paradox of Economic Backwardness and Health Development (Thiruvanathapuram, 1984).
5 This portion draws substantially from C. P. Mathew and M. M. Thomas, The Indian Churches of St. Thomas (Delhi, 1967); L. W. Brown, The Indian Christians of Saint Thomas (Delhi, 1990); James, Demographic Response To Economic Crisis: A Study of Syrian Catholics in Kerala in Historical Perspective (New Delhi, 1995)
6 Robin Jeffrey, The Decline of Nair Dominance: Society and Politics in Travancore 1847-1908 (Delhi, 1976); T. C. Varghese, Agrarian Change and Economic Consequences: Land Tenures in Kerala 1850-1960 (Bombay, 1970); P. K. Michael Tharakan, Migration of Farmers from Travancore to Malbar, from 1930 to 1960: An Analysis of Its Economic Causes (Thiruvanathapuram, 1976); Census of India, Travancore (Thiruvanathapuram, 1875), I.
7 Census of India, Travancore (Thiruvanathapuram, 1931), I.
8 Mathew and Thomas, Indian Churches, 21-30.
9 Genevieve Lemercinier, Religion and Ideology in Kerala (Delhi, 1994).
10 Rajan, Bhat, and Dyson, Fertility and Mortality Transition in Kerala: A Study Based on the Parish Records of Kerala (Thiruvanathapuram, 1993).
11 James, Demographic Response, II; Rajan, Bhat, and Dyson, Mortality and Fertility Transition in Kerala: A Historical Investigation (Delhi, 1998); James and Rajan, "Short Run Variation," 57; Rajan and James, "Indian Parish Records: A Qualitative Assessment," Annales De Demographic Historique (forthcoming); Wrigley, "Birth and Baptisms: The Use of Anglican Baptism Registers as a Source of Information about the Number of Births in England before the Beginning of Civil Registration," Population Studies, XXXI (1977), 281-312.
12 Wrigley and Schofield, Population History, 66-86, adopted the same method for the analysis of seasonality in England's parishes.
13 Lee, "Short-term Variation: Vital Rates, Prices, and Weather," in Wrigley and Schofield (eds.), Population History; idem, "Demographic Response," 5; Galloway, "Basic Patterns," 281.
14 The previous literature available in this context established that a four-year lag is long enough for understanding the effect of an exogenous or endogenous variable on the dependent variable in short-run demographic analysis. See Lee, "Short-term Variation," 356-401; Galloway, "Basic Patterns," 282; Galloway, Secular Changes, 85-93.
15 Seiver, "Trends," 93; Udry and Morris, "Seasonality," 675; Wrigley and Schofield, Population History, 33-65; Rajan, Catholics in Bombay: A Historical, Demographic Study of the Roman Catholic Population in Bombay (Calcutta, 1993).
16 Rajan, "Infant and Child Mortality among Roman Catholics in Bombay," Journal of Family Welfare, XXXVII (1991), 50-61; idem, "Marriage and Remarriage among Bombay Roman Catholics," Journal of Family Welfare, XXXVI, (1990), 61-79; Wrigley and Schofield, Population History, 66-86.
17 Seiver, "Trends," 95; Udry and Morris, "Seasonality," 678; Lee, "Methods and Models," 337-370; Richards, "Weather," 208; Rajan and Sandhya Rao, "Can We Explain Demography through Culture?" Man in India, LXXI (1991), 383-399; Rajan, "Marriage," 65.
18 Wrigley and Schofield, Population History, 66-86.
19 Rajan, Catholics, 65-121.
20 Rajan, "Demographic Transition among Roman Catholics of Bombay: An Examination through Parish Records," in Krishnamurthy Srinivasan and Kunju Bihari Pathak (eds.), Dynamics of Population and Family Welfare (Bombay, 1991).
21 Lee, "Short-term Variation"; Richards, "Weather," 210
22 Richards, "Weather," 207; Lee, "Demographic Response," 5.
23 Wrigley and Schofield, Population History, 66-86; Richards, "Weather," 210.
24 Richards, "Weather," 210; Lee, "Short-term Variation," 356-401.
25 Rajan, "Marriage," 65; Bhat and Shiva S. Halli, "Demography of Brideprice and Dowry: Causes and Consequences of Indian Marriage Squeeze," Population Studies, LIII (1999), 129-148.
26 Wrigley and Schofield, Population History, 66-86; Richards, "Weather," 210.
27 Thomas Robert Malthus, An Essay on the Principles of Population (London, 1985; orig. pub. 1798).
28 Lee, "Short-term Variation," 356-401.
