Johnston] 



NAVAHO POPULATION 179 



by the Navaho, it then follows that the lower medians obtained from 

 reported Navaho age distributions reflect a fairly persistent under- 

 count (or underregistration) of the adult Navaho population." 



The discrepancy between the summary characteristics of Models 

 E through / (whose parameters attempt to approximate more recent 

 fertility and mortality conditions among the Navaho) and those of 

 the Navaho age distributions since 1930 is even more marked. The 

 majority of the Navaho medians in this period, excepting those derived 

 from the 1939 and 1957 rolls, fall between 16 and 17 years, while those 

 of Models E through / fall between 18 and 20 years. Here also, it 

 might be concluded that the reported Navaho age distributions reflect 

 an imdercount of adult Navahos. Such an interpretation would also 

 serve to explain why most of the Navaho child- woman ratios (given in 

 table 32) are considerably higher than those derived from most of the 

 later models, despite the fact that the respective crude birth rates are 

 generally similar.^^ 



Model J merits special comment in view of the extremely high rate 

 of natural increase (4 percent per year) which it reflects. The implied 

 crude birth rate (49) for this model is very nearly identical with the 

 rate reported for a closely observed area of the Navajo Keservation 

 in the 1955-59 period (see footnote 3, p. 152 f .) . Furthermore, both the 

 median ages and the child- woman ratios derived from this model are 

 remarkably close to those reported for the Navaho area and the Navajo 

 Eeservation populations in the 1960 census. It would appear that 

 Model /, which was originally selected to illustrate the maximum con- 

 ceivable rate of natural growth, is a good representation of the current 

 Navaho area population. If so, this area will continue to experience 

 a very rapid population growth in the future. 



It should be stressed, however, that such similarities may be 

 fortuitous. The 10 models shown in table 36 are merely representative 

 of a large number of alternative models which might be constructed 

 on the basis of slightly different values and combinations of values 

 assigned to the basic parameters. These models are valuable heuristi- 

 cally, insofar as they provide indicators of the values which might 

 be typically expected under given conditions of fertility and mortality. 

 However, the use of values derived from these models to "predict" 

 or "adjust" the data obtained from the Navaho or other populations 

 does not appear to be warranted. 



" It must be recognized, of course, that heavy mortality immediately before and during 

 the Fort Sumner period, followed by a rapid recovery, would produce a "younger" popula- 

 tion. Furthermore, the mortality levels typified in the models may not apply in the 

 Navaho situation. 



IS The reader will recall that the average child-woman ratio among the more representa- 

 tive Navaho age distributions for the 1920-57 period was 760. The corresponding average 

 for Models E through H is only 6S7. The average crude birth rate implied in these same 

 models, however, is 39.1 per 1,000, which is very close to the rate reported among the 

 Navaho in 1956 and 1957. 



