¹ The first quotation is from Cushner, Nicholas P., Spain in the Philippines: From Conquest to Revolution (Quezon City: Ateneo de Manila Univ. Press, 1971), p. 196Google Scholar; the second is found in Gironiere, Paul de la, Adventures of a Frenchman in the Philippines (1857; rpt. Manila: Burke-Miailhe Publications, 1972), p. 3.Google Scholar

² There is scant but provocative evidence of these local mortality trends in published sources describing the colony as a whole, e.g., fragmentary data for the 1876–1898 period published in the census of 1903. U.S. Bureau of the Census, Census of the Philippine Islands (Washington, D.C.: Office of the Census, 1905), III, 11–17.Google Scholar

³ On the causes of the transition from a mercantile to a self-sufficient economy, on the nature of the new programs for development, and on the ramifications of the transition for Philippine society, a number of useful sources are available. For general descriptions see, for example: Cushner, Spain in the Philippines, chap. 8; Costa, Horacio de la, “The Formative Century, 1760–1860” and “Philippine Economic Devel opment,” in Asia and the Philippines: Collected Historical Papers, ed. Costa, Horacio de la (Manila: Solidaridad Publishing House, 1967)Google Scholar; Stanley, Peter W., A Nation in the Making: The Philippines and the United States, 1899–1921 (Cambridge, Mass.: Harvard Univ. Press, 1974)CrossRefGoogle Scholar, chap. 2. For specific areas, see Larkin, John A., The Pampangans: Colonial Society in a Philippine Province (Berkeley: Univ. of California Press, 1972), chaps. 3, 4Google Scholar; and McLennan, Marshall, “Land and Tenancy in the Central Luzon Plain,” Philippine Studies, 17, No. 4 (1969), 651–82Google Scholar. Maria Lourdes Diaz-Trechuelo covers a range of issues in her series of five articles in Philippine Studies (1963–1965).

⁴ See de la Costa, “Philippine Economic Development,” pp. 130–41.

⁵ McLennan, “Land and Tenancy in the Central Luzon Plain.”

⁶ Norman G. Owen, “Kabikolan in the Nineteenth Century: Socioeconomic Change in the Provincial Philippines,” Diss., Univ. of Michigan, 1976.

⁷ The monopoly was organized in 1782 and remained in force until 1882. Population growth occurred in these areas mainly as a result of state programs to expand the land area devoted to tobacco cultivation. Cushner, Spain in the Philippines, pp. 202–4.

⁸ See White, Benjamin, “Demand for Labor and Population Growth in Colonial Java,” Human Ecology, I, No. 3 (1973), 217–36CrossRefGoogle Scholar, for a useful discussion of the impact of cash cropping upon peasant fertility. See also the comments by Clifford Geertz and Etienne van de Walle, and White's reply two issues later.

⁹ Smith, Peter C., “Changing Patterns of Nuptiality,” in A Demographic Path to Modernity: Patterns of Early-Transition in the Philippines, ed. Flieger, Wilhelm and Smith, Peter C. (Quezon City: Univ. of the Philippines Press, 1975).Google Scholar

¹⁰ See, for example, Constantino, Renato, The Philippines: A Past Revisited (Quezon City: Tala Publishing Service, 1975).Google Scholar

¹¹ For discussion of some of these issues see Larkin, John, “The Place of Local History in Philippine Historiography,” Journal of Southeast Asian History, 8, No. 2 (1967), 306–17.CrossRefGoogle Scholar

¹² The parish inventory project is described in Peter C. Smith, “A Continuing Inventory of Primary Materials in the Parish Archives of the Philippines: Five Years in the Field,” paper presented to the Social Science History Association Workshop on The Sources of Asian History and the Generation of Quantifiable Historical Indicators, Toronto, Canada, 28–29 February 1976, which also summarizes the results through October 197;. Some 14 provinces had been inventoried by that time.

¹³ In addition to the primary data from individual parishes, there are unexplored data on population change in the archives of the various religious orders that worked in the Philippines and in the Philippine National Archive (PNA). The Manila Archdiocesan Archive (MAA) contains certificates of cemetery interment which I have not used because they are incomplete and pertain mainly to the Ciudad de Manila and nearby parishes.

¹⁴ The parish data from Cebu Island are now undergoing intensive examination at the East-West Population Institute in Honolulu. For a preliminary report see Cullinane, Michael and Smith, Peter C., “The Population of Cebu Province, Philippines, in the Nineteenth Century: A Working Paper in Historical Demography,” paper prepared for the 1977 annual meeting of the Midwest Conference on Asian Affairs,De Kalb,Illinois,14–15 October 1977Google Scholar

¹⁵ We have in hand aggregative results from 50 parishes, but have so far been unable to obtain population estimates for all fifty. For some parishes, the pattern of baptism, burial, and marriage frequencies suggests inaccuracies or discontinuities in the series.

¹⁶ Robles, Eliodoro G., The Philippines in the Nineteenth Century (Quezon City: Malaya Books Inc., 1969), pp. 77ff.Google Scholar; and Michael Cullinane, “The Spanish System of Demographic and Spiritual Accounting,” unpublished.

¹⁷ I have used tally forms similar to the ‘Preliminary Extraction’ forms developed by the Cambridge Group for the History of Population and Social Structure. See Wrigley, E.A., ed., An Introduction to English Historical Demography (London: Weidenfeld and Nicolson, 1966), pp. 113–15.Google Scholar

¹⁸ In most cases the gaps involve a month or two; in other instances, one or two years or somewhat longer. Short gaps generally result from lost or damaged pages in the record, especially the first or last few pages of a volume when bindings have given way and covers detached. Larger gaps result when signatures or whole volumes are missing.

¹⁹ In each case, the average of the nearest ten valid (not imputed) monthly frequencies before and after was substituted. Of the 40 parishes examined in this study, 22 required no imputation and 7 required imputation for gaps of only one month at most. Eleven parishes had gaps of from 2 to 11 months, and only 13 had gaps of greater than one year. (These frequencies are not additive because a few parishes had gaps of more than one type.) The longest gap, ten years (120 months), occurred in four parishes.

²⁰ The imputation procedure is a conservative one with respect to estimating the prevalence of crisis mortality: I assume in effect that no mortality crises occurred during gaps in the record, but it might be argued that the gaps index periods of social or economic disruption, perhaps including high mortality.

²¹ These documents have been made available to me through the kindness of Cardinal Archbishop Jaime Sin, Bishop of Manila. The documents entitled Plan de Almas, Manila Archdiocesan Archive, contain annual records of population and vital events drawn from the Estado de Almas compiled annually in the parishes.

²² Non-Christian Chinese constituted a very small proportion of the population of individual parishes; Christian and non-Christian Chinese combined numbered some 40,000 or 0.9 percent of the colony-wide population in 1870 (figure cited in Wickberg, Edgar, The Chinese in Philippine Life 1850–1898 [New Haven: Yale Univ. Press, 1965]. p. 58.)Google Scholar

²³ This difficulty stems from the variable fashion in which annual parish counts were submitted to, or collected by, central authorities. (In the original parish padron or plan de almas, the reference dates are entirely unambiguous.) The records in the archdiocesan archive today are found in dated bundles (e.g., “1787–1804”), and each folio within them is dated by single year but not by month. Sometimes it is not clear whether the date on a folio is that of completion of all the parish censuses listed, or of receipt in Manila of all of these, or of entry to the consolidated record. Often a folio will include data for the 20 or so parishes of a province, all described by a single date. It is likely, however, that the individual parish counts were completed at different times throughout the year or even the year before. There is also the possibility, suggested by evidence in a few cases, that parish counts which had not been received in Manila by the time of compilation were replaced by those for the preceding year.

²⁴ By fitting a straight line through the data points in log form, we define a curve with increasing absolute growth (decline) over time and a constant rate of change.

²⁵ In his analysis of 54 English parishes, Schofield uses a ratio of 2 to 1 with the average of 20 surrounding years as base. See Schofield, R. S., “Crisis Mortality,” Local Population Studies, 9 (Autumn 1972), pp. 10–22Google Scholar. D. Turner takes 1½ to 1 as his criterion for “minor crises” with 2 to 1 defining “major crises,” in “Crisis Mortality in Nine Sussex Parishes,” Local Population Studies, 11 (1973), pp. 40–42Google Scholar. M.W. Flinn looks at the percentage excess of mortality over normal (in the ten years surrounding the crisis) and calls this his Crisis Mortality Ratio. A 30 percent excess is regarded as a crisis; see Flinn, , “The Stabilization of Mortality in Pre-Industrial Western Europe,” Journal of European Economic History, 3 (1974), 285–318Google Scholar. Lastly, Del Panta and Livi Bacci regard (under specified conditions) a 50 percent increase over normal deaths as a small crisis, and a multiplication of normal deaths by four as a major one; see Panta, L. Del and Bacci, M. Livi, “Chronologie, Intensité et Diffusion Des Crises De Mortalité En Italie: 1600–1850,” Population, 32 (special number, September 1977), 401–46.CrossRefGoogle Scholar

²⁶ Hollingsworth, T.H., “A Formula for Assessing the Relative Severity of Population Crises,” International Journal of Environmental Studies, 7 (1975), 119–22.CrossRefGoogle Scholar

²⁷ Briefly, he argues that “intensity” must incorporate objective and subjective components. The objective event (how many died) is reflected in (_nq _j) ‘(_nP _j) and inversely related to the number of survivors, (_nP _j) (1–_nq_j). “Suddenness” is indexed by 1/n, more sudden crises being more intense (harder to bear for those involved). Subjective stress is related to the numbers of people involved in an episode, _nP _j, discounted to _nP _j⅔ because a large population is not comprehended by an individual in the same fashion as a small population. Duration is discounted to n^1/3 to take into account the diminished psychological absorption of “others” by an individual over time. We should also note that Hollingsworth takes the arbitrary level I ≥ 20 to signify a mortality crisis, whereas we are dealing with far less catastrophic episodes (I ≥ 1.0).

²⁸ Minja Choe, the East-West Population Institute Data Analysis Officer, has developed a packaged program, ‘CRISIS,’ to perform these calculations.

²⁹ The two-month index is either 2Ij or 2Ij-1 depending on whether data for the preceding or the following month are used.

³⁰ Hollingsworth has pointed out in a personal communication that space (parishes) as well as time might be aggregated (preserving contiguity) in order to find the largest possible indexes. Unfortunately, we cannot pursue this useful suggestion with our small sample of areas.

³¹ It also affords some chance for cross-national comparison since Hollingsworth's index was used in several papers for a recent conference on mor tality crises. See the proceedings of the IUSSP International Colloquium of Demographic His tory, Montreal, Canada, 8–10 October 1975. It would probably be instructive to test several different crisis indexes through simultaneous use, but we have not yet done this. As our data base grows, we will be able to explore the effectiveness of alternative indexes in a range of situations.

³² Hollingsworth, T.H., Historical Demography (Ithaca, New York: Cornell Univ. Press, 1969)Google Scholar; Durand, John D., “The Population Statistics of China, A.Ḋ. 2–1953,” Population Studies, 13, 3 (1960), 209–56Google Scholar; McNeill, William H., Plagues and Peoples (New York: Anchor Press, 1976).Google Scholar

³³ R.S. Schofield, “Crisis Mortality.”

³⁴ Klein, Ira, “Death in India, 1871–1921,” JAS, 32(1973). 639–59.CrossRefGoogle Scholar

³⁵ Unfortunately, we do not have burial data from the Ciudad de Cebu itself.

³⁶ Madridejos, Bishop Benito Romero de, Pastorales Demos Disposiciones Circuladas a los Parrocos de esta Diocesis de Cebu, 2 vols. (Manila: Establecimiento Tipografico del Colegio de Santo Tomas, 1883).Google Scholar

³⁷ For further contrasts to the Cebu data see the excellent description of mortality patterns on the island of Samar, especially in the period of the 1880s, in Robert Bruce Cruikshank, “A History of Samar Island, the Philippines, 1768–1898,” Diss. Univ. of Wisconsin—Madison 1975. Cruikshank examines the impact of Moro raids on food supply (p. 6$ff.) and thus on mortality patterns.

³⁸ Stamp, Dudley L., The Geography of Life and Death (London: Fontana, 1964)Google Scholar; and De, S.N., Cholera: Its Pathology and Pathogenesis (Edinburgh and London: Oliver and Boyd, 1961).Google Scholar

³⁹ Worchester, Dean C., A History of Asiatic Cholera in the Philippine Islands (Manila: Bureau of Printing, 1909).Google Scholar

⁴⁰ Surely the most infamous is the cholera epidemic which originated at the port of Manila in September 1820 and led to the October 9th massacre. Paul P. de la Gironiere, Adventures of a Frenchman, p. 31F.

⁴¹ In a personal communication Norman G. Owen has provided some interesting information on an apparent diffusion sequence in the Bicol region. During the 1882 cholera crisis, the peak number of deaths in the pueblo of Nueva Caceres came in September; similarly for Polangui. In Guinobatan one peak occurred in October and another in March of the following year (coinciding with the peak of deaths in Camalig). Philippine National Archives, Erection de Pueblos, Camarines Sur, II, “Pesquisa de Camarines Sur”. Owen sees this as “tentative evidence of the epidemic passing down the central valley of Kabikolan (NE to SW).”

⁴² S.N. De, Cholera.

⁴³ Worchester, Asiatic Cholera.

⁴⁴ On the Ilocos Coast some 90 percent of the annual rainfall occurs during this period. Wernstedt, Frederick L. and Spencer, Joseph E., The Philippine Island World: A Physical, Cultural and Regional Geography (Berkeley: Univ. of California Press, 1967), pp. 609–10.Google Scholar

⁴⁵ Whereas deaths fell below or rose above the average monthly tally by an annual average of 7 percent in 1970, monthly variations in our parishes in the 1850s (a period relatively free of crisis) average 18 percent. The 1970 figure was calculated from official data on registered deaths. See Bureau of Census and Statistics, Vital Statistics Report 1970 (Manila: Bureau of Census and Statistics, 1973)Google Scholar. The figure for the 1880s is the combined pattern for the 40 parishes over the 10 years.

⁴⁶ Fernandez, Benito Legarda y, “The Philippine Economy Under Spanish Rule,” Solidarity, 2, No. 10 (1967), 1–21.Google Scholar

⁴⁷ Adas, Michael, The Burma Delta: Economic Development and Social Change on an Asian Rice Frontier, 1852–1941 (Madison: Univ. of Wisconsin Press, 1974)Google Scholar; and White, “Demand for Labor and Population Growth on Colonial Java.”

⁴⁸ Smith, Thomas C., The Agrarian Origins of Modern Japan (Stanford: Stanford Univ. Press, 1959)Google Scholar; and Dore, R.P., “Agricultural Improvement in Japan 1870–1900,” Economic Development and Cultural Change, 9, No. 1, pt. 2 (1960).CrossRefGoogle Scholar

⁴⁹ Land ownership rapidly concentrated in elite or Chinese mestizo hands, beginning a trend which continues to the present. In addition the large estates of the religious orders were largely put into leasehold agriculture. See Wickberg, The Chinese in Philippine Life 1850–1898, and McLennan, “Land and Tenancy in the Central Luzon Plain.”

⁵⁰ Lewis, Henry T., llocano Rice Farmers: A Comparative Study of Two Philippine Barrios (Honolulu: Univ. of Hawaii Press, 1971).Google Scholar

⁵¹ These events were preceded in the record by reference to a “smallpox epidemic” in July-December 1876, simultaneous with “drought” and “poor harvest” that year, and were followed by comments on a “major fire in poblacion” in April 1879. Unfortunately, we cannot determine the impact that these local events may have had on mortality, since the burial records of the period have not survived.

⁵² Eliodoro G. Robles, The Philippines in the Nineteenth Century, pp. 32ff., 229ff. Robles cites a contemporary observer who claimed that the Central Insitute of Vaccination had only 127 vaccinators at the end of 1897.