HomeMy WebLinkAboutAPA829J
fIB
3525
1\.4
845
c.2
Migration and Population Dist-
ribution in Alaska:1974-1990
by Daniel A/Seiver
March 1,1976
~/1 II A .d-·r t-f,I ?Q
(
r
'-....._-
.-
MIGRATION AND POPULATION DISTRIBUTION
IN ALASKA:1974 -1990
,...
..
By Daniel A.Seiver.
Institute'of Social,Economic,
and Government Research
Anchorage,Alaska
March 1,1976
Population Study Report Pursuant
to Contract _
with
Federal-State Land-Use Planning Commission
,',
(
\
"MIGRATION AND POPULATION DISTRIBUTION
IN ALASKA:1974 -1990 11
1.DETERMINANTS OF INTERSTATE MIGRATION AND ANALYSIS OF PROJECTED HIGRATIOW
I.A.Presentation of r·1AP Data and Analysis on Interstate Migration ...·.
I.A.1.Introduction
Many studies of interstate migration in the United States,and
studies of interregional migration in other countries,show that employ-
ment opportunities ~nd interr.egiona1 income differentials are important
determinants of migration decisions.'Many other variables have been
shown to b~.significant in these studies,but most empirical \'lOrk has
,
utilized cross-section data.These data are useful for hypothesis test-
,ing,but'not for projection.
Alaska time-series data are available for net civilian migration for
the years 1960-1974,and also for average annual total .civilian employment.
The growth of civilian employment {EMPGRO}is a good proxy for employment
opportunities.As a proxy for income differentials (\'lhich include wage
differentials),a measure of income in Alaska relative to the United States
was constructed.This measure is the ratio of real disposable personal
income per capita in Alaska to the same quantity in the United States.
(RELINC).2 In addition to reflecting real wage differentials,RELINe
also reflects the relative tax structure and individual economic well-
being (per capita personal income).
l ,
"..---"(-
r---'(-,
\
- 2 -
Using these three time-series on net civi 1ian migration,civi 1ian
employment growth,and relative income,multiple regression analysis
enables us to estimate the effects of employment growth and relative
income on net migration.Figure 1 shows both the historical time-
series for net civilian migration and the series predicted by regression
equation (1).The "fit"is fairly close.Details of the estimating pro-
cedure are contained in the Technical Appendix to this report.The es-
timating equation is:
~0 0IJI~o ~,t
(1)NE~MIGt =-;~.~57 ~1.~36 x EMPGRO t +57.557 x RELINC t _1 -3.541 x DUMMY
where
NETMIG~=Net civilian migration to Alaska,in year t •..
EMPGRO t =Et -Et_l"'w~ere Et =total civilian employment in
Alaska in year t,
RELINC t _1 =Ratio of Alaska real disposable personal income per
capita to U.S.real disposable personal income per
capita 1n year t-l,
DUMMY =1 in 1964,for earthquake.
I.A.2.Employment growth and net migration
Equation (1)shows the independent effect of employment growth on
net migration.Other things equal,growth of 1,000 in civilian employment
will lead to net civilian migration of 1,236.This figure is quite reason-
~
able given that many workers migrate to Alaska with few or no dependents,
T,,,
(~\
- 3 -
and that the resident labor force will supply some of the additional
labor demanded.3 An earlier study of migration to Alaska also showed a
very close link between'employment growth and net migration.4 The large
increases in employment projected by the MAP models thus have a potent
effect on net migration~and thus population.
I.A.3.Relative income and net migration
Employment growth in low-wage industries,coupled with rising taxes
and a rapidly growing population~is likely to lead to less migration than
a regime of employment growth in high-wage industries,with falling taxes....
and a slowly growing population.While these are clearly extreme cases~
relative income trends can have a reinforcing or offsetting effect on the
.'
amount of migration induced by a given amount of employment growth.The
relative income proxy used in equation (1)takes account of all of the
above factors and has a significant independent effect on net migration.
The 1972 value of the RELINC term was about 0.85.An increase to 0.86,
other things equal,lmplies net migration in the following year of 570.
The total effect of the RELINC term is more corpplex.Since an increase in
Alaska population will reduce RELINC,the additional migration this year
will increase population and tend to reduce RELINC the follm'/ing year,re-
ducing net migration.This servomechanism effect is substantial for large
amounts of migration,which are projected in some years of the scenarios
discussed below.The relative income ratio is lagged one year to allow
t}me for information to filter down to the rest of the'United States.
- 4 -
,~It is hypothesized that the functioning of the labor market in the
state makes the response to employment growth much quicker.5
I.A~4.The age-sex distribution of interstate migrants
(
,
Much of interstate migration in the United States is attributable
to males and females aged 20-34.6 Migration to and from Alaska"in the
past as well as the future,has and no doubt will conform to this pattern.
Census data for Alaska for the 1965-1970 period 7 s~o~a pattern of sub-
stantial inmigration in the 20-24 age groups,and persistent though small-
er,net outmigration ·in every age group above 35-39 for both sexes.Few
Alaskans'retire in Alaska.The under-40 data is highly suspect,however,
since military migration is included in the totals and cannot be controlled
for.S The fol1owing·sche~e.was used to determine the age-sex composition
of the projected net migration streams:for age groups :above 35-39,the
net outmigration rates for the 1965-1970 period v/ere calculated.These
rates,reported in Table 1,are applied to the specific age-sex categories
in each year of the projection period.For the under-40 population,the
J965-1970 pattern of net migration to California from non-contiguous states
was used.These "migration percentages"add up to 100 and are also re-
ported in Table 1.For example,if net inmigration ina given year is
10,000,2,590 of the migrants will be males aged 20-24.To determine
total net inmigration in the under-40 group,the amount of net outmigra-
~tion in the over-40 group is calculated.To this quantity is added net
migration calculated from equation (1).This total equals total net
t.
- 5 -
Table 1
MIGRATION RATES (40+)OR PERCENTS (0-39)
"--.---<--Age Group Male Female
0-1
1-4 2.4 2.8
5-9 1.7 2.0
,...
10-14 .2.4 2.8
15-19 10.4 4.9
('"19.020-24 .'25.9
25-29 8.2 10.0
30-34 2.5 1.8
35-39 1.1 2.0
40-44 -0.16 -0.22
45-49 -0.16 -0.18
50-54 -0.28 -0.18
55-59 -0.26 -0.26
60-64 -0.68 -0.84
65 +-0.82 -0.78
".
- 6
r'\inmigration,which is then allocated as noted above.Thus,by definition,
(2).NET}1IG =MIGIN -MIGOUT
and by adding MIGOUT to both sides,
(3)NETMIG+MIGOUT =MIGIN.
(
Thus,given EMPGRO,RELINC,and the values in Table 1,it is pos-
sible to project net migration,and also the age distribution of the
migration.EMPGRO is supplied directly by the NAP model.9 The ~umera
tor of RELINC is 'also supplied by the MAP model,and for the projections
made below,the U.S.denominator is expected to grow at 1 percent per
year through 1989.10
I.B.Analysis of Future Interstate Migration Streams Projected
Under Alternative Development Scenarios
I.B.l.Introduction to MAP scenarios
The MAP petroleum scenarios are described i.n detail elsewhere ll and
only a short summary will be presented here.The scenarios are important
because the pace at which Alaska's petroleum resources are developed is
likely to be the primary determinant of the state's economic growth~The
.current formulation of the model provides three alternative scenarios for
the period 1974-1990:(1)a "limited development"case~in which present
developments are carried forward (essentially Prudhoe Bay),a few additional
r I
~-
I '.r
\
- 7 -
fields are opened near existing areas and the federal OCS leasing pro-
gram is limited to the Gulf of Alaska.Total oil production reaches
2 million barrels a day by 1980 and 4 million barrels a day by 1990.12
(2)An "accelerated development"case in which,in addition to case (1),
new petroleum areas are opened up in the northwest,both onshore and off-
shore,and a second North Slope Oil pipeline is constructed,mainly as
a result of leasing in Naval Petroleum Reserve No.4.In this case,oil
production reaches 7.7 million barrels a day in 1990.(3)A "max imum
development"scenario which approximates the maximum rate of petroleum
development that could oc~ur -in Alaska.13 -The rate of development is
.
comparable to that envisioned in the plan in IIproject Independence."It
is assumed.in this scenario that in addition to case (2),the Federal gov-
ernment leases heavily in the Bering and Chukchi Seas.This,in turn,
necessitates construction of oil and gas pipelines running from north to
south in western Alaska.Availability of the pipelines and processing
facilities would then make additional leasing feasible in the new western
areas for Native corporations and the state.Alaska's oil production reaches
5.2 million barrels a day in 1985 and nearly 10 million barrels in 1990.
The other key variable determining the size of the state's oil reve-
nues (and the pace of economic growth)will be the price of oil.The
model has been run for three cases:wellhead prices of $7,$5,and $3 per
barrel in 1973 prices (equal to $11,$9,and $7 per barrel market prices).
These nine (3 policies x 3 prices)cases are analyzed beloww 14
·I
(
- 8 -
the M7 case to very moderate amounts of net migration,comparable to
;
(
- 9 -
pre-pipeline years,in the L3 case.These variations over time are a
direct result of the variations in employment growth projected by the
MAP model in the different scenarios.There are several years,however,
when the relation M>A>L for net migration is partially reversed.A case
in point is 1984-85 in the L5-A5-M5 comparison of Figure 3.For these
~wo years,emplo~ment growth is greater in L5 than A5,which is enough
to make L5 net migration higher than A5.Other "reversals"in Figures
2 - 4 almost always reflect temporary reversals in the pace of employment
growth.
Another factor accounting for the "saw-tooth ll patterns of projected
"..
net migration is the operation of the RELINC variable noted above.A
year of very high net migration leads to rapid population growth ·which
in the following year tends to reduce RELINC,and.thus net migration.If
employment growth slackens also,there will be an even sharper change in
net migration.
I.B.3.Age-sex pattern of migration and the age-sex distribution of
the state1s population
Section I.A.3.described the methodology used to allocate net migra-
tion to Alaska to age-sex categories.Given a 1974 age-sex distribution
for the state15 ,the effects of different development scenarios (at $5 oil)
on the age-sex distribution in 1990 are shown in Figures 6 (A and B)and
7 (A and B).Figure 6 (A)shows the 1974 age-sex distribution for reference.
_/
-10 -
The major differences are not surprisingly in the age-sexcate-
gories most prone to migrate.The 20-34 year olds,male and female,
account for the bulk of the increase in the M5 versus AS case,skewing
a youthful 1974 age distribution even more toward these youthful ages.
The abso1ute and relative numbers of children and teenagers-is also
higher,partly as a result of net migration of families with young child-
ren,and partly because the bulk of female fertility occurs during the
years 20-34.The opposite effects are noted in the L5 case as compared
to the AS case.Thus,there is a direct relation between the pace of
economic.development.and the ·comparativeyouthfulness of the population,
operating through the parameters determining the allocation of net mjgra-
tion to age-sex categories.This effect is a corollary of the direct
link between net migration,and thus population,and the pace of economic
development,which was analyzed above.In the sections below,the effects
on total .populati on of the state and the population of the reg'ions of the
state are analyzed in detail.
II.STATEWIDE AND REGIONAL POPULATION GROWTH:PROJECTIONS AND INTERPRETATION
II.A.Detailed Analysis of State and Regional Population Projections
Under Alternative Scenarios
II.A.l.State population growth
Figure 8 shows the projected state population for 1974-1990 for
the three development scenarios at $5 oil.The sharp variations in net
"
/.-.-....
I \
I "
-11 -
migration do not appear,since 1)in anyone year,net migration is
only a small fraction of total population,and 2)the population grows
each year by.the excess of births over deaths (natural increase)which
offsets net outmigration,and adds a stable element to·population growth.
The cumulative effect of net migration 'is large,however,since the varia-
tions in 1990 population are a direct and indirect result of variations
in net migration.State population in M5 of 874,000 is 40 percent greater
than the L5 population of 620,000.Figure 9 shows the effect of varia-
tions in price of oil on state population in the accelerated case.Higher
oil prices lead to·substantia"lly,higher populations'~with A7 population
.
exceeding A3 by 160,000 or 25 percent,a somewhat smaller difference than
the range between MS and LS.
In all cases,population grows in all years except 1977 or 1978,\,then
.negative net migration exceeds natural increase.In all·other years,the
two sources of population growth reinforce each other,:since the age-sex
distributi'on of net migration tends to increase fertility;;:
II.A.2.Regional population growth:introduction
A 1970 'Census map of Alaska with regional delineations of the MAP
model is reproduced in Figure 10.The regions generally correspond to
the regio~s delineated by Rogers i~earlier work on Alaska 16 ,with ~air
banks separated from the Interior region and Anchorage separated from the
Southcentral region,giving seven regions in total.
-12 -
Full-scale regional population models have not yet been constructed,
and thus regional populations are determined by an allocation process
discussed in detail in the Technical Appendix.The allocation process
essentially distributes regional population on the basi~of regional ~n
p10yment growth.
II.A.3.Population growth and employment growth in urban regions
The Anchorage region,.a1ready the largest and most urbanized region
of the state in 1974,grows absolutely and relatively in the 1974-1990
projection period.'"Anchorage'l s population in the AS case reaches 390,000
in 1990,'as shown i~,Figure 11.It is also clear that the MS and L5 cases
have.strong effects on An<:horage's population,with MS 1990 population
100,000 above AS,and LS 1990 population about 60,000 less than AS.Even
in the LS case,however,Anchoragels population doubles from its current
level.This rapid growth is a direct result of rapid employment growth.
Figure 12 depicts employment growth in Anchorage in the A5 case.From
a base of 72,000 in 1974~employment more than doubles to 191,000 by 1990.
Table 2 shows the effects of rapid population growth in Anchorage on its
share of the state's total population in the'LS,AS,and M5 cases.The
percentage rises substantially in all three cases but is highest in M5,
reaching 56.3 percent by 1990.Some further implications of this growth
are discussed in Section II.C.
"
(
-13 -
The Fairbanks region and the Southeast region contain the remain-
ing urbanized areas of the state.Population growth in these regions
is depicted in Figure 13,and their respective shares of total population
are listed in Table 2.While both Southeast and Fairbanks grow fairly
rapidly,neither grows nearly as fast as the Anchorage region,and
Table 2 shows that each loses ground to Anchorage throughout tbe pro~
jection period.17 In the MS case,their combined population percentage
drops from 29.4 in 1974 to 22.3.EmpJoyment growth for these two re-
gions is,of course,slower than Anchorage's.Figure 12 shows that even
though employment in both Fairbanks and Southeast grows fairly rapidly
....0 •
after 1918,neither region can match Anchorage's pace.In"absolute terms,
however,the populations of both Fairbanks and Southeast grow substantially,
reaching 84,000 and 87,000,respectively,in 1990.
II.A.4.Population growth and employment growth in rural regions
The columns labeled "all other"in Table 2 show that the regions \'/hich
comprise about 90 percen~of the state's land area will have a declining
share of total state population in the projection period,falling from 27.1
percent to 22.3 -21.4 percent in 1990.Most of this decline occurs after
1983;in fact,in the MS case,the 1983 share is higher than the 1974
share.Once again,the trend in employment growth (Figure 12)provides
the explanation.Employment rises rapidly to a peak in 1983 and by 1990
has not returned to the 1983 level.This employment pattern in the AS
case reflects additional pipeline construction and oes development which
(
!
Table 2.",
,.'i .
,'REGIONAL PERCENTAGES OF STATE POPULATION 1974-1990
!\i:
I
1974
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
,',
L5 CASE
'1
A5 CAS'E .M5 CASE
",'
#5 #7 #3 Other #5 #7 #3 Other #5 #7 #3 Other
.
13.943.5 15.5 13.9 27.1 43.5 15.5 27.1 43.5 15.5 13.9 27.1
43.2 15.3 13.7 27.8 43.2 15.3 13.7 27.8 43.2 15.3 13.7 27.8
44.2 14.8 13.9 .27.1 44.2 14.8 13.~27.1 44.2 14.8 13.9 27.1
45.3 14.5 14.0 26.2 45.5 14.4 14.1·26.0 45.5 14.4 14.1 26.0
45.5 14.4 13.9 26.2 45.4 14.2 13.7 26.7 45.4 14.2 13.7 26.7
46.1 14.2 13.9 25.8 45.9 14.0 ,13.7 26.4 45.9 14.0 13.7 26.4
46.5 13.8 13.7 26.0 46.4 13.5 13.4 27.7 46.4 13.5 13.4 26.7
47.0 13.6 13.6 25.8 46.5 .13.3 '13.0 27.2 46.6 13.3 13.0 27.1
47.7 13.4 13.5 25.4 47.1 13.1 12.8 27.0 46.9 12.8 12.6 27.7
48.0 13.1 13.3 25.6 47.8 12.7 12.7 26.8 47.4 12.2 12.3 28.1
48.7 12.9 13.2 25.2 49.1 12.6 12.9 25.4 49.0 12.1 12.5 26.4
49.5 12.8 13.2 24.5 50.0!'12.5 .12.9 24.6 50.4 12.1 12.5 25.0
50.3 12.7 13.1 i 23.9 :i'50.9 I 12.4 '12.8 '23.9 '!50.6 11.6 12.1 25.7
50.9 12.6 12.9 23.6 51.7 12.2 12.6 23.5 51.3 11.2 11.7 25.8
51.7 12.4 12.7 23.2 52.6 12.0 12.4 23.0 53.8 11.2 12.7 22.3
52.5 12.2 12.5 22.8 53.4 11.8 12.2 22.6 55.0 11.1 11.6 22.3
53.2 12.1 ',12.4 22.3 54.2 11.6 '12.0 22.2 56.3 10.:9 i 11.4 21.4
I .
;I :;i i
"I '.
-15 -
impacts these regions,but the employment growth is essentially exo-
genous and temporary.In fact,Region 4,the Southcentra1 region,is
growing fairly fast throughout the projection period,and by 1990 in
the A5 case,has a population (95,000)exceeding both Fairbanks and
Southeast.In 1990,then,Southcentral should probably be considered
an urbanized region.If Southcentra1 is removed from the "all other ll
.
category,the share.of population in the "all other"category drops
markedly.Its 1990 population share drops to just 9.0 percent.
Figures 14 and 15 show population and employment trends with the
regions aggregated according.to Rogers'scheme:Southcentral is combined.."...
with Anchorage (the Greater Anchorage Region)and Fairbanks is combined
with Interior.Nor'thwest and Southwest are combined for the lIother"
category.Two-thirds of the state's population resides in the Greater
~ncho~age Region by 1990,and this region has almost two-thirds of the
state1s total employment by 1990.Regions 1 and 2,comprising more than
half of the state's land area,contain only 7.6 percent of the-state's
population.The L5 and,M5 cases do not produce dramatically different
results with respect to population shares.But in the M5 case,the popu-
lation share of the Greater Anchorage Region rises to 69 percent of the
total in 1990,and the share of Northwest and Southwest combined falls a
little further to 7.4 percent.
(
/
-16 -
II.B.Intrastate Mobility and Relative Employment Growth
II.B.l.1965-1970 pattern
The 1965-1970 relationship between employment growth and intrastate
mobility can be traced out with some difficulty,given the limited amount
of 1970 Census mobility data.18 Table 3 shows employment growth for each
region in the 1965-1970 period,and the number of intercounty migrants
1965-1970,living in each region in 1970,after deductions for armed
.forces and college student mobility.19 The analysis is complicated by
the fact that Anchorage is the only region comprising one census division
(county).and thus .fQr.all .the.o'ther regions,'the number of intercounty
(inter-census division)migrants .is an overstatement of the number of
interregional migrants.Nonetheless,the pattern of rapid employment
growth in Anchorage dra'llfng.intrastate migrants is clear.Only the North-
west reglon has a higher rate of employment growth,reflecting the Prudhoe
Bay discovery in 1968.Of the total employment growth in the state,An-
chorage accounts for more than half.Southeast,Southcentral (Kenai-Cook
Inlet oil)and Fairbanks.account for most of the remaining employment
.growth (38 percent)•.The mobility pattern discernible from the Census
.data in Table 3 shows a similar pattern,keeping in mind that all regions
except Anchorage have overstated totals of·interregional migrants for the
reason noted above.In addition,we cannot calculate net intrastate mi-
gration even between counties,since there is no data on the origin of
these intercounty migrants.For example,all net intrastate migration
might have been to Anchorage,with all other regions simply trading migrants
(i
-17 -
Table 3
EMPLOYMENT GROWTH
AND INTERCOUNTY MIGRANTS 1965-1970
EMPLOYMENT GROWTH
1965-1970 1970 RESIDENTS LIVING
IN A DIFFERENT ALASKA ._---."-_._-_._-"
---.---_._-----REGION Quantity Percent COUNTY IN 1965
--_._~---_._---
1 1,141 45.0 771
2 1,001 26.4 1,050,...
3 2,848 20.5 2,912
/'4 .2,289-"32.1 3,909(
'.
5 11 ,317 36.9 4,411
6 * *
310
7 2,943 25.6 1,968
Sources:Mobility -u.S.Census,QQ.cit.in footnote 19
Employment -MAP historical data bank
*=less than 0
,---_/
,/----....,
(
(
-18 -
among themselves.With this additional caution in mind,we can still
note that 63 percent of this intercounty migrant pool resided in just
five counties in 1970:Anchorage,Fairbanks,Juneau,and r'latanuska-Sus-
itna and Kenai-Cook Inlet.This distribution follows the employment
growth distribution fairly close1y.20
II.B.2.Projected pattern
Without a set of explicit regional population models,it is impossible
to measure precisely the pattern of intrastate migration implied by the MAP
population projections.~t i~possible,however,to assign approximate
.!..
values to the rate of natural inc~ease in each region for the years 1974 -
1990.Then,by comparing actual projected 1990 regional population witho
the 1990 regional populatlon which would result from natural increase
alone,it is possible to estimate the combined effects of interstate and
intrastate migration on regional population for comparison\oJithrelative
regional employment growth.Table 4 lists the 1974 and._J.99_0.populations
of the seven regions in the AS case,and also the 1990 regional populations
resulting from 1.7 percent per year natural increase alone.21 Relative
employment growth for the projection period is also shown.The Southwest
and Interior regions actually grow more slowly than would beimp1ied by
natural increase alone;that is,the balance of net migration between the
region and the rest of the state and the United States is negative.North-
west's net migration is slightly positive.These three regions have very
"
-19 -
Table 4
1990 REGIONAL POPULATIONS AND EMPLOYMENT GROWTH
A5 CASE
PERCENT OF
TOTAL STATE 1990 POPULATioN
EMPLOYMENT GROWTH 1990 POPULATION WITH 1.7%RATE
..REGION 1974-1990 A-5 OF GRm.JTH
1 3.0 19,090 18,023
~,
2 "·4.6 35,363 36,775
3 ~2.8 86,769 65,051
,
(/
4 9.3 94,761 60,599
5 60.6 390,255 203,027
')
-~
6 1.2 10,103 11,421
7-8.6 83,865 72,181
-,
;;
i
(
\
-20 -
sma 11 shares of state employment gro\'lth.The great bulk of the excess
of total projected regional population over the natural increase popula-
tion is accounted for by Anchorage.The Anchorage region is almost twice
as large in 1990 in the AS case as it would be if it only grew by natural
increase.The Southeast,Southcentral,and Fairbanks regions also grow
by migration,although Southcentral actually has as much implied migra-
tion as Southeast and Fairbanks combined.The projected relative employ-
ment growth over the 1974-1990 period do~s not fit the pattern of popula-
tion growth precisely,reflecting different "elasticities"of population
growth with respect to employment growth.22
.-...
II.C •.Implications'for Population Distribution
-The sections above make clear that in 1990 most of the land area of
the state of Alaska will be almost as empty as it is today even though
Alaska's population will have grown quite rapidly.It is also clear that
the pace of economic development will have an important effect on the
rapidity of population growth.But,in all cases,over half of the
state's population will be located in the Anchorage region,in a large
city even by United States standards.Anchorage's "suburbs",part of the
Southcentral region,will also be growing rapidly.Fairbanks will increase
its relative importance in the northern regions of Alaska,but Alaska will
truly become a one-city state.Much of Anchorage's growth will result
from interstate migration to Alaska but,in addition,much of the intra-
state migration will be directed to Anchorage.The highly 'Skewed geographic
(
/
\
"/
-21 -
population distribution will no doubt have far reaching economic and
non-economic consequences.The age-sex distribution will also be highly
skewed toward youth,making Anchorage the recipient of much interstate
migration,a particularly "young II city,which could also have important
economic and non-economic consequences.The identification and analysis
of these consequences are beyond the scope of this paper.One interesting
possibility,however,is that individuals with different life-styles will
be able to co-exist in Alaska:the urban life-style of the East and-West
coasts of the United States will be available in Anchorage and its sub-
urbs (with a few adjustments .for climate)\'/hile the distinctly Alaskan..'...
life-style,which requires much land area per person or family,will-still
be attainable in Alaska's still relatively uninhabited hinterlands...-
i
it
t.
f
tt
i
~
!
i•;
...,
)
I
I -
/
I
V
,.
,...."
Figure 1
ACTUAL (--.--)AND PREDICTED (---)
NET CIVILIAN MIGRATION TO ALASKA,
1960 ;.1974
"'\,
\
\
\
-...
1-t
i
i6-1
I
~r,
s-f
t•I
1
4-'i
!;.
3-i
I
i
I2-;:
i,
;
••!
i
o-i
!
i•
1 -;
I
!
i I
_ 2 -t
J.-.-.-"---r--~--r-r..,.........--r~I--l~'60~6~·~r~3 62-64 63-65 64-66 65-67 66-68 67-69 68-70 69-71 70-72 71-73 72-74
- 3
11 ~
t•
10 -i.i
9.JI. .j
8~;·to .J
\~
"~TilOUS.}-22 -
17 -:'---:--"!"-r--.
I
I
t16-';
./~,~\.,-t f
t ;/
1 (
f /.'14 ~1 1
13 i I'i )lc
12 -'·I•
,.,'.
(Tho~s.).r-------------------,..'--+--_
35 _
.
-23
t -
f
1
f
1
i
!.,
~
I
i
d
1,
~
I'.•,
·i
i
J;·:.,
\·i
{
,
i
f,":.
I;.,
·
..
A3
(
Figure 2
.,
NET CIVILIAN MIGRATION 1974-1990
OIL PRICE =$3.00
,
25 ...J
\t•,
t
20 j
,.
-10...:..-:·.;·
-15 _-;:;·,
i I i I I I
85 86 87 88 89 90
i I Ii'I767778798081
I I
,Q7l1.75
''';';-20 ~"~
i
i
t
(ThOUs..)..•-24 -
'_·---·...5.-,;,;,;,._.-._-.-.
I,
10 -:
I
I
5
o
- 5
i 10
MS
---AS
LS
Figure 3
NET CIVILIAN ~IGRATION 1974-1990
.OIL PRICE =$5.00
·--~'l
i,
J
t
I
I
i
I
1
I
I•!
1,
I
I
i
I
l
{.,
f
l"
I
I
~
i
t
I
I
\
I'
i
i
i
(,
!
i
1
i·I
1
I
i,
i
~
f
I
•\•:•!
-15--19-7""~-7-r15--r716--71:---7T""j8-7"-9-"'-'8~O-8TI1-8"T12-""813-~81r-4-8T'"15-8rI6-aTI7-8""a-8""19~'""':910-._..---_..--.....
!
I
i,
I
J
I
Ir
)
I
I·•i
tr
I
I
I
~
i
It;
·.:;
i
"
-~O -
f47
A7
Figure 4
NET CIVILIAN MIGRATION 1974-1990
OIL PRICE =$7.00
n
\
\
\
\
\
\
I
--,
\.
"
M7,
A7,
L7
i
I,..,i
I
0--':
!
i
I·i
5 l-I
1
!
.5~
I
25 -i
f
t('--l
~O \
I
I
I
15 ~,•·,•r
1
1
10 --.1.I
I
i
I
~(Thou:s.)
. .r----.-..---.--'--------.----..---..."..,...,...--.-.....---------,-,._~."--"'-
5"0 1
I
/-,1
(5'-1
I,
)
40~I
I
I
35 ~t
l
I.~
)
J •i30--;;
{
t
J
201
.!i
,
~.I I I I I I I I I I I ,I r J I I
1974 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90
i
i
\
!
i
1
l
i
i
t,
f
i,
t
I
i
I
!
!
!
I
I
I,
!
i
f·:
~,
i·
..
A7
AS
-26 --
...------....---.......--..------·---~·-~·..···-··...---·1
I
I
I
f
I
t
1
t
l
I.I
I
l
~
i
i,•t
i
I
!
i
~
i
i
i
.i
j
j
1
~
Figure S
NET CIVILIAN MIGRATION 1974-1990
ACCELERATED DEVELOPMENT
".
35 .-l~:---------.,.-_.__.__.....o'
j
I
I
I
(~.I
;\:;JO-!
i
'j
i
i
I
j25
i
115-;
i
I
I
I
20 ..JI
)
tA7,
lAS,
1.5 JA3i
{
f
1.0 .J
t
i
I
°1'._-......,.~
f r
/i t
. i 5 \..-,/-I
l
A.'1974
)
,"
Male Female
7
.,.
8
Figure 6
ALASKA AGE-SEX DISTRIBUTION
B.1990 -L5
""
65 +
60-64
55-59
50-54
45-49
40-44
35-39
30-34,
25-29
20-24 f
15-19
10-14
5-9
1-4
"'--'-:""0-1 '
I
8 7
Male Female
J
6
i
7
J
S
A.1990 -AS
,,'
Male Female
Figure 7
ALASKA AGE-SEX DISTRIBUTION
B.1990 -M5'.Male Female
-----..--.-.-.-'-_._-'-----~--'------_.'--.._-,.._..._--_._~
\•t
f
I
I
1,
~
J
1
'j
I
t4S
AS
LS
Figure 8
.'
.'.
,.~.
PROJECTED STATE POPULATION,1974-1990,
(with oil =$5.00)
-.
;.
l••~r
~
~
1~
600,000!
I
\
/----.....
(:50,J'-,"-../1
f
~
1
'~
;·
,400,000-1
'l~ooo~doo -29 -.l---.'--._---...----:-----.-----;------...,.---...-...-----..-f
(-,I
900,...,Ju-,
I
t
i
!
800,000 -I
i
I
·lt
700,000 1
,
1.·-l300,000 I
i,
j
;
~.
200,000 ....;
•1
I
1
i
t
I
lq''0-1
'-.-
o I , I I I I ( I , I I ...(..+,.i .I I.,'v,...,C::'7~77 7R 7Q Rn ~l A?R3 alf.85 86 87 88 89 90
I
I
I•!
i
)
:r
~
l
I
(
!<!
;
!
f
I
1
~
i
-f,
'-•I
}
t
j,
~
~
~
i
>··t
t·,
r-1
'I
1
i
~·
·,\
i,
~
ic•;s•t
!.
5
-30 -
,...
PROJECTED STATE POPULATION,1974-1990,
WITH ACCELERATED DEVELOPMENT AND THREE OIL PRICES
{
\
l
~-,
..
T,000,000 _'.~r-'------------.-·--~------~--·---··----l·
9~~>~Figure g I
j •
t
r
I
l
!•;
i
t
i
~
I
i
•i
j·l;
800,000~
!
i
i••2
.700'0001
)600,000-1
i
I
J
)5(\r~,,0-f
\,(,I
-'"-_/if
i
f
(
,1
::400,000 i
i·I-i '~300,000-1 .
j -
i
-I,,
,200,000---1,
1
c
I
10:~OO 1
i
\
-----0:":---..!../~-=----'Ir--~l -"T'I---='I~-rl -"T'j--rj---rj-"'jr----rj--TI-'I-l,r--rl-Tj-.,.j-Ilr-----.i
197~75 7677 78 79 80 81 82 83 8~85 86 87 88 89 90
5 "I
,-,.,~
1
~,.
3 4 6 7 .,8 9 10 }c .~--..
.(,/."-....r.,...·-'~.,,I t •
\
)
"
A
Figure 10
A.,.,ALASKA REGIONS
'I '"Northwest''~.'.
;!0 1 ,
"I 'Inter'ior•
B 8 1 '0
6
I,
\"0 \,.
c o.C
.
-
'.
'...0 So~thwest
0 .'"'''-'''''''0 w I~~.-.:....._-_.-.....
2 I I
,' 0 I·
IE 0 .........._--.......:..........---_.--_........"
SCAI.E E i'.,I,1If!.J s;o........~n·l~....2!fO WILts.,..~,.'.
~
~I J
I~.""'~..~II I
",South\oJest (part).\,I t
-<1 ,.d)c:,-().'"(
F
<fi •0 I F'I..()~~.,!JD?!"8.,""I, .'0'..~"..o 0-I It·p '.•
o t't:I .-..0
a
8 3 4 5 6 7 8 9 19 I
Source:State r.1ap in U.S.Bureau of the'Census,.Q.E..,ci t.,pp •3-81..1
.
f
J«
j
$
~c•5
!.
j
i
I
AS
LS
I I I I I
86 87 a8 89 90
...-------.--..-..----•.----.------,,~----.---J
H5 I
!
I
I
i.
i
-32 -
I I ,
80 81 82
I
79
,
78
..
Figure 11
PROJECTED POPULATION OF ANCHORAGE
1974-1990,('t/ith oil price =$5.00)-
I I , I
197".75 76 77
.(
..........,.../
SOD ,O'CO .+
I
~
400,000
··:i .
i
l·J
100,0001
I
f,
!
j
"I
!
~t
f
;300,OOO---t
~
~. J
J
~
~J .;.
,200,000~{
i·.~
J
I
~
-33 -
~E:mi"lo~-~-----------_..-...-..~=,.....",..,.....,....:-=-.....-----_•._.-._""~,,,-.•_,.---••---.•.._..•.-------.~~-.,
~(\.
\Figure 12
REGIONAL EMPLOYMENT 1974 -1990
A5 CASE
,
i
L
J
!
·.i
i•S
t
i
i
i
•~
.~...
..
Regions
-----------------
-,
50,000
r --~·"'OO
""
200,000·-
10,000 ---~--1-9-7""~--7"'15-7-r16--7'7--7Ir8-7'~--Sr b-a'll--aTf2·--'SI3-·-ST"14--·-'81r-5-8'16~~alr-7-S'~--8rI9-9"O-------'
,
i
i
I
,0
I
I.
!
?
1
i
t
t.'
f
f
I~
~
'.
."'-...-..o--=-'-r
I
I
I
0,
~....-'+~..o.o ..·
--....--
1974 -1990
..
AS CASE
Al~Other Regions
~,
lr
l
~r•i·•
~......_--.
1Region
\.
150,000 i
~f .
/
/..
-34 -
1t
€
(
~
~
1
~~
l•1
i
i,
~
i
~
I
---i'.-.'
f
I
~
~
I
f
!
!
j
.j
i
!
~
1~..
;..
.=
f
~
)
;f
;,
1;
i
i
i
j
i
I10,000 ---=---·-jr--"'lr-"'I--r.-'.-'Tj-""'j-orI-rl---'jr-i/-ij-••-.Tj-TI-rl----;Ir--o
--'
,a~u ?~7h 77 78 79 80 81 82 83 84 85 86 87 88 89 90
$-,:nnag -r~-------------··o,-,,-----------------
j Figure 13
(/~'0 f REGIONAL POPULATIONS
\~
I~
o i
\;
~
!.i:
!
I
J
o tor
---l200,000 ~
(
C-tt
f~-t
(0Ci1
.'\'.,>1.
,.,
S~i:li10g ,
(
\-".,.-'
400,000
300,000
.200,000
r
!
1
I,
I
I
I
I
]
35 -.._
Figure 14
POPULATIONS OF ROGERS'REGIONS
1974-1990
AS.CASE
Regions 4 and S
....
..
1(.~"fJO
......--.
·50,000
Regions 7 and 6
~
.Regton 3
Other
.-----------~_.
~-.
~
.~---....----...-.!._-
I I;
-·10J~~~--~-r,-T,-~,~~,-~,-~,~-,-~~~~--~-~-~~_~-__~III I ,.,14 I I I I I I'O?U 75 76 77 78 7g an 91 n?A~n 95 An 97 SA p,q qn
Figure 15
EMPLOYMENT IN ROGERS'REGIONS
1974 -·1990
AS CASE
------~'.,....,..·····l-·
f
I
J
f
'1
!•:
t.
--..-~
-----
36 -----._-_.,--------,..."........,.-
/.~
"
....Jr..,
\S~rdHg9-y---------------'
-\
I,
I
i
~,~
\--'1
1
1
1ti
.;j
~
~
I.i
200,000 -,.J
;I
I
f
--i
.1
I
1
f
~
it.50,000 -gf .
~
~.
~Regions~7 and 6~
5
--~
::.,r._
:Reglon 3
i)
$,
!Other,
~
100(
'-,
10,000 ---::----1-:Q...,7~.--·-7....ls-...,71,-6-7-:7--7.,.la--7T""~-8...,~--8.,...ll--,81r-2-8..,.13--8'T1~--8r'5-·aI6~--a.,...17--8T""la-...,019--96----·
-37 -
FOOTNOTES
1.R.Paul Shaw,Migration Theory and Fact (Philadelphia,1975),and
Michael J.Greenwood,"Research on Internal Migration in the
United States:A Survey,"Journal of Economic Literature (June,
1975),pp.397-433,provide comprehensive bibliographies of the
migration literature.
(
2.Amore sophisticated denominator would take into account the effects
of distance on information flows and the propensity to migrate,
thus using a U.S.average,weighted by distance from Alaska,of
the major regions of the U.S.
3.This local supply will vary directly with the unemployment rate and
the.skill lev.~1 .of t~e population.
4.Ar10n Tussing,et~.al.,Alaska Pipeline Report (Fairbanks,1971),'p.11.
-...-
5.The relative income term has another equilibrating effect:if employ~
ment growth does not call forth sufficient labor supply (partly sup-
plied by migration),then the real wage schedule will tend to rise
faster,raising the RELINC term and inducing more migration.
6.larry long~"New Estimates of Migration Expectancy in the Uni'ted -
States,"Journal of the American Statistical Association (March,1973),
pp.37-43..
7.U.S.Bureau of the Census,Census of Population:1970,Mobility for
..States and the Nation (Washington,1973).
8.Thus,the aggregate Alaska table shows net outmigration in the 25-29
age category,which is unacceptable on a priori grounds.
9.For details of the MAP model,see David Kresge,"Projections of !\laska1s
Growth to 1990,11 Alaska Review of Business and Economic Conditions,
forthcoming.
"
(
\,
(
-38 -
10.This rate exceeds the negative rate for the period 1973-1975,and
falls short of the rate achieved in the 1960-1968 period in the
United States.Both of these periods are perhaps abnormal for
the United States economy.The 10-year 1966-1975 growth rate is
about 1.5 percent,\-Jhich,if substituted in the model,makes little
difference in the projections •
.11.Kresge,10c.cit.in fn.9.
12.Projections of oil production and employment are based on Thomas More-
house,"The Future of Alaskan Petroleum Development,"unpublished
manuscript.
13.This scenario is dependent upon optimistic assumptions about economically
recoverable reserves and the availability of the capital and techno-
logy necessary·to develop these reserves.
14.The AS (Accelerated Development -$5jbarrel oil)case has been chasen
as the "base case"for this analysis and thus has been analyzed in
-much greater detail •.-The Technical Appendix discusses other special
assumptions made for these computer runs of the MAP regional model.
15.This age-sex distribution is an estimate.based on the actual 1970 age-
sex distribution of the civilian,non-Native,non-military dependent
population.It is assumed that the numbers and age structure of the
military and military dependent population are unchanged over the
projection period.The Native population is assumed to grow at 2.0
percent per year,the recent growth rate.By 1990,the civilian non-
Native non-military-dependent population comprises about 80 percent
of the total.
16.George W.Rogers,Alaska's Population and Economy (College,1963),vol.I.
17.The effects of a capital move from the Southeast region to the South-
central region have not been modeled.The capital move will clearly
result in a much lower 1990 population for Southeast and a much
higher one for Southcentral.Anchorage1s population would also
be increased.
..
-39 -
18.A good deal of caution must be used in analyzing the results,however,
since the Census data on mobility have a number of weaknesses,and
the Census data must be compared to employment grolt/th calculations
based on Alaska's Department of Labor data which is not strictly
comparable to Census data.
19.These calculations are based on Table 119 of General Sodal and
Economic Characteristics,U.S.Census of Population:19zo,
Alaska,PC(l)-C3.
20.The mobility data is based on residence in 1965,so many migration
streams are not measured,for instance,an interstate migrant
arriving after 1965 and moving intrastate in the 1965-1970 period.
Multiple moves (including return moves)are also not measurable.
The distribution of net interstate migration cannot be measured
precisely either•
.....
21-
"-?(
22.
Rates of natural increase do not vary that much by region._The rate
used (1.7 percent)is an estimate of the Alaska rate of natural
increase for 1975.
The calculation of population-employment elasticities is discussed in
detail in the Technical Appendix •
•
·.',
"
<
\.TECHNICAL APPENDIX
I.Special Assumptions*
The behavior of Native regional corporations has not been explicitly
modeled.For the computer runs discussed in this report,it was assumed
that the only link between state personal disposable income and Alaska
Native Fund payments would be the 10 percent minimum dividend to be paid
to shareholders.The exogenous income (DNCS)for the years 1974-1990 is
presented in Table Al.The assumed schedule of exogenous local government.....
borrowing fRBOLX)is also shown in Table AI.This series declines to zero
as local government revenues begin exceeding expenditures in the early
(1980's~
\
II.The Migration Equation
The estimate of net civil i an migration to the state in any one year
can be subject to substantial error.To reduce the effects of errors in
maasurement,b~o-year moving sums of the net migration,employment,and
relative income series were constructed which reduce the a~mount of erratic
fluctuation in the dependent variable,at the cost of one degree of freedom.
*These assumptions resulted from consultation with Mr.Paul Engleman
of F.S.L.U.P.C.
"~
./~,
-A.2 -
Table Al
DNCS and RBOLX
YEAR DNCS (r~i 11 i on $)RBOLX (Million $)
..!:L A3,M3 L5 .A5,M5 L7,A7,M7
1974 8.1 8.1 8.1 8.1 8.1 60
75 8.5 8.5 8.5 8.5 8.5 65
76 7.9 7.9 7.9 7.9 7.9 45
77 2.1 2.1 "...2.i 2.1 2.1 30
78 2.5 2.5 2.6 2.7 2.8 20
79 3.0 3.0
3.2 3.3 3.5 10
80 3.4 -·3.5 3.9 4.0 4.4 0
(81 3.9 4.0 4~6 .4.7 5.2 0"82 4.3 4.5 5.2 5.3 6.1 0
83 4.7 4.8 5.7 5.9 6.5 0
84 5.0 5.2 6.3 6.5 6.5 0
85 5.3 5.5 6.5 6.5 6.5 0
85 5.6 6.1 6.5 6.5 6.5 0
87 6.0 6.5 6..5 6.5 6.5 0
88 5.3 6.5 6.5 6.5 6.5 0
89 6.5 6.5 6.5 6.5 6.5 0
90 6.5 6.5 6.5 6.5 6.5 0
-A.3 -
Figure 1 graphs the two-year sum variable against its predicted value.
The complete statistics for the migration equation are reported below:
(Al)NETMIG t =-52.557 +1.236 x EMPGRO t +57.557 x RELINC t _1 -3.541 x DUMMY
(S.E.)(0.114)(18.234)(1.128)
F(3,9)=55.0
/;.
Each coefficient in (Al)is highly significant,as is the entire set of
(coeffitients.Most of the variance·in the dependent variable is nexplained"
{'''-
by the independent variables.
Sources of the data are as follows:net migration -annual estimates
of Alaska Department of Labor;employment -MAP data bank;relative in-
come -Alaska-Neville Beharie,"Alaska Disposable Personal Income,"ISEGR
mimeo,Alaska relative price index from fiJAP data bank,and population from
Department of Labor.United States -Economic Report of the President,
February,1975,Tables C-18 and C-45.
"'.
.,.,
-A.4 -
III.Regional Allocation of Population
A series of 10glinear regressions of total population on total employ-
ment,one for each region,was estimated for the period 1965 -1973.For
estimation purposes,regions 1,2,and 6 were combined,with the constants
in all cases adjusted to fit the final historical observation.The equa-
tions are listed,by region,below:
(A2)REGION 1:POP =2.1344 +0.3360 x ENP
(S.E.)(.028)
2:POP.·::-2.5203 +0.3360 x EMP
(S.E.)(~028)
3*:POP-=1.5438 +0.7508 x EMP
4:POP =1.3176 +0.9185 x EMP
(S.E.)(0.153)'
5:POP =1.1389 +0.9236 x EMP
(S.E.)(.034)
6:POP =1.5723 +0.3360 x EMP
(S.E.)
7:POP =1.5903 +0.7595 x EMP
(S.E.)(.152)
POP =Population
EMP =Employment
R2 =.9284
R2 =.9284
R2 =.7668
R2 =.9852
R2 =.9284
R2 =.6940
*An average e1astfcity based on Census data \'las substituted in the
Southeast region.
"t.t.
-A.5 -
Given that the equations are loglinear,the coefficients of the
employment variables represent elasticities.These vary in accordance
with expectations,with the Anchorage and Southcentral regions approaching
"1.0,Fairbanks and Southeast somewhat lower,and the other regions much
lower.
In each year of the projection period,employment is calculated
for each region,as is total state population.The employment figures are
"plugged inll to equations (A2)to give first-round estimates .of regional
population;the sum of.these regional estimates is also calculated.Each
regional population estimate is then divided by the sum of the regional
population estimates to give the Ilpopulation percentage ll for each region
implied by equations (A2).These percentages are then multiplied by the
state tota-l population,giving the regional populations.~This process is
summarized in equationA3:
(A3)
.f(EMP t r)'x '~f{EMP t r)
r '
/"..
"\,
\...-/
where POPt =region r population in year t,r
STPOP t =state population in year t
f ( )=population-employment equations (A2)