Page 35 - IJPS-11-1

International Journal of
Population Studies Indian Census Rolls: Data for U.S. tribes

and other summary indicators, along with the Excel code Both the Lummi and non-Lummi populations have
that generates this information from the data in worksheet; more males than females (Table 2). There are 5% more
(7) Census Data, a worksheet containing the transcribed males than females in the Lummi population, but the sex
data from the digitized microfiche file, M 595, which were ratio is much more skewed for the non-Lummi population
entered independently by different people on different (1.59). The much higher sex ratio is influenced by the
spreadsheets and reconciled (where there were differences) relatively small number of records for the non-Lummi
into a single worksheet; and (8) Coding Inconsistencies, population roll. This table illustrates that small numbers
which lists the records (by microfiche page roll) found to may be an issue when analyzing individual tribal data from
have inconsistent coding between Tribal identification and Indian census roll records, especially when subdividing the
Ward status. The majority of these inconsistencies are due data by demographic and tribal characteristics.
to the fact that children whose parents are clearly Lummi As noted, Indian Census Rolls also contain data on
members have not yet been enrolled.
marital status. Table 3 shows the population broken down
Although not shown in Appendix A1, the Indian by marital status (married, single, and widowed) and tribal
census rolls indicate people who died after January 1, 1937, identification. The small number issue is also apparent in
by placing a line through them and indicating the date of this table, especially for non-Lummis, which have only 5
death. The last two columns (AC, and AD, respectively) in and 6 persons single or widowed, respectively. For the total
worksheet 7 display this information. One column (AC) and the Lummi populations, most are single (59% total and
indicates if a death has occurred, and the second column 63% Lummi). Around 30% of the Lummi population is
(AD) lists the date of death. Nineteen deaths are recorded married compared to 83% of the non-Lummi population.
in the Lummi census roll. On the first row of every new In both populations, between 7% and 8% of the population
household, the template accumulates data at the household are widowed. Due to relatively small counts, caution is
level enabling the calculation of total household population, needed when interpreting the non-Lummi data.
the total household population by blood quantum, and
persons per household for each household both on and off Table 4 and Figure 1 show the population in 5-year
the reservation. The total household population by blood age groups by gender for the Lummi population. The
quantum is calculated by summing up the blood quantum age and gender breakdowns were not large enough to be

fractions for each person in the household. For example, meaningful for the non-Lummi population.Moreover, the
if the three persons in a household are 50% Lummi, the age data were missing for 43 non-Lummi males and 20
household population would be 1.5 for both the Lummi non-Lummi females. The shape of the population pyramid
and non-Lummi populations. (Figure 1) with its broad base and narrow tapering at the
older ages is indicative of a population with high birth and
3. Lummi housing and demographics, 1937 death rates (Britannica, 2023). Also notable, there are more
males (36) than females (22) above the age of 60 (Table 4).
There were 731 people counted in the 1937 Lummi Indian This pattern suggests that male survival rates were higher
census roll. All but one lived in households and that person was than female survival rates at older ages 90 or so years
listed as a resident of San Quentin Prison in California (that
is, this person resided in “group quarters,” not a household).
The household population of 730 lived in 194 households. Table 1. Lummi household characteristics, 1937
The majority were enrolled in the Lummi tribe 661 (661 Households Household Persons per
of 731 or 90%) while 70 persons (10%) were not enrolled population household
(N.E). Table 1 shows the Lummi household population On Lummi Reservation 157 598 3.81
living on the reservation was 4.5 times larger than those Off Lummi Reservation 37 132 3.57
living off the reservation (598 vs 132). Correspondingly, the Total 194 730 3.76
on-reservation Lummi population lived in 157 households
and the off-reservation population in 37 households. As a Source: National Archives (1965).
result, the average number of Lummi living in households
was somewhat higher (3.8) on the reservation than off the Table 2. Lummi population by gender, 1937
reservation (3.6). Larger household sizes and overcrowding Total Male Female Sex ratio
on Indian Reservations are long-standing issues (Pindus
et al., 2017). In 2018, for example, the average household Lummi 661 338 323 1.05
size was 3.6 on Indian Reservations, which is higher than Non-Lummi 70 43 27 1.59
the average for the general U.S. population, which was 2.6 Total 731 381 350 1.09
(Rodriguez-Lonebear et al., 2020). Source: National Archives (1965).

Volume 11 Issue 1 (2025) 29 https://doi.org/10.36922/ijps.3906