Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Introduction
The City of Boston is in the midst of a remarkable growth period, and Mayor Walsh is seeking to chart a
new course for the city through the development of Bostons first master plan in over 50 years: Imagine
Boston 2030. Previous MAPC demographic projections,1 published in 2014, indicated that the citys
population would reach 709,000 people by 2030, requiring construction of over 53,000 housing units.
These forecasts served as the basis for the citys 2014 housing plan2 and remain the official projections
used for transportation planning and other regional or state efforts.
The citys continued rapid growth since 2014 prompted the Boston Planning and Development Agency and
MAPC to reexamine MAPCs existing projections to determine if an updated set of forecasts could better
serve the Imagine Boston 2030 (IB2030) process. Over a six month period, staff from the BPDA Research
Division and MAPC Data Services Department collaboratively reviewed recent trends in births, deaths,
migration, household formation, and vacancy rates and worked together to prepare a new set of
projections. BPDA staff constructed a new population model for the city to produce new projections of
population by age at five-year intervals out to the year 2030. The methods and results of that model are
described in a separate BPDA document: Methodology for Bostons Population Projections 2016.3 MAPC
staff used those projections as inputs to our existing housing demand model (with some minor adjustments)
to produce updated projections of households and housing demand for the city out to the year 2030.
This document summarizes MAPCs work in four parts: the first section describes key similarities and
differences between MAPCs 2014 population projections and those produced by the BPDA; the second
section describes the key assumptions of the household/housing demand model and the production of
alternative forecast scenarios; the third section describes the results of the analysis and MAPCs
recommendation for updated housing production targets; and the fourth section describes preliminary
estimates of population and households in 2050. After producing a range of forecasts using different
assumptions about household formation rates, vacancy rates, and dorm construction, MAPC recommends the
use of a mid-range projection of 58,400 housing unit needed between 2010 and 2030. The full data
products describing this update can be found at data.mapc.org/projects.
It should be noted that while MAPC is producing these supplementary projections in collaboration with
BPDA to support the IB2030 process, these household estimates do not supersede the projections released
in 2014, which remain the agencys official projections for Metro Boston and its municipalities. MAPC will
conduct a wholesale review and update of those projections beginning in early 2017, and will incorporate,
to the extent practical, the information compiled by the BPDA to inform its population model.
Comparison of MAPC and BPDA Population Projections
The BPDA population projection model varies from MAPCs methods in a number of important ways. Most
notably, the BPDA model uses more recent fertility and mortality statistics than were used by MAPC and
estimates gross domestic and international migration at the municipal level (rather than net migration).
Despite these differences, the projected 20-year growth in total population is not drastically different for
the two models: 14.8% according to MAPCs 2014 forecasts and 17.1% according to the new BPDA
projections. The total population of the two projections varies by only 2% in 2030, a difference of 14,000
people.
http://www.mapc.org/projections
https://www.boston.gov/finance/housing-changing-city-boston-2030
3 http://www.bostonplans.org/research-maps/research/research-publications
1
2
Page 1
While the total population for 2030 is similar across the two projections, the projected age structure does
differ somewhat. Both sets of projections anticipate that the number of 20-24 year old residents will fall,
and expect the number of 30 39 year olds to grow considerably. However, the BPDA projections for
these age ranges are higher than MAPC, by about 5%, due principally to assumptions about migration
rates for the post-Millennial generation. Conversely, the BPDA projections anticipate somewhat less
robust growth in the 40 54 year old age groups, with the 2030 population about 5% smaller than
MAPCs projections. Partly as a result of this smaller cohort, the under-20 population is about 1.2% smaller
than in the MAPC projections. Both sets of projections anticipate that the under-15 population will rebound
from recent declines, with growth rates of 14% (BPDA) and 16% (MAPC). Similarly, both agencies project
large increases in the senior population (>65), though MAPCs projections are slightly higher for the 65-79
year olds, and BPDA shows substantially larger growth for residents 80+ years old. Overall, the BPDA
projection for all residents 65+ is 5.4% higher than MAPCs.
100,000
90,000
2000
80,000
2010
70,000
MAPC 2020
60,000
BPDA 2020
50,000
MAPC 2030
40,000
BPDA 2030
30,000
20,000
10,000
0
Age Group
As demonstrated below, the differences in the two population projections has important implications for the
housing demand forecasts. More 20-39 year olds will result in a greater number of nonfamily households,
in particular, and the larger senior population anticipated by BPDA will mean more elderly householders
and/or assisted living/nursing home residents.
Page 2
0-4
5-9
10-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75-79
80-84
85+
45.0%
40.0%
35.0%
30.0%
25.0%
20.0%
15.0%
10.0%
5.0%
0.0%
50.0%
40.0%
30.0%
20.0%
10.0%
0.0%
Census 1990
Census 2000
Census 1990
Census 2000
Census 2010
ACS 2010-2014
Census 2010
ACS 2010-2014
The dorm construction goal and progress toward it are described in the report City of Boston Student Housing
Trends: 2015-2016 Academic Year, produced by the Department of Neighborhood Development, accessible at:
https://www.boston.gov/sites/default/files/boston-student-housing-trends-ay15-16.pdf. Dorms containing 2,400
beds were completed between 2011 and 2014, with an additional 6,000 net beds in the pipeline.
4
Page 3
living alone.) As shown in the charts on the previous page, family headship rate trends have generally
declined since 1990 for age groups under 35 and over 54. (35 54 year old rates dropped from 1990
to 2000 but them remained steady. Meanwhile, non-family headship rates had been increasing steadily
since 1990 for most age groups, but saw a sharp reversal after 2010 for all age groups under 54. We
attribute this to the citys high rental housing prices which are discouraging may residents from forming
their own households. Whether this is an anomaly due to the current housing market or a long term trend
likely to continue can only be analyzed over time. Given this uncertainty, MAPC produced two scenarios of
demand based on different assumptions about non-family headship rates (both scenarios use the same
2010 family household headship rate.) The low headship scenario carries forward the 2014 non-family
headship rates, while the 2010 headship scenarios assumes 2010 headship rates for all ages and
categories.
Household demand was further disaggregates into attributes such as household size, income, housing unit
type and tenure using PUMS 2010-14 data. Income categories are based on those defined by the US
Department of Housing and Urban Development (HUD), which characterize households both by their size
and their income relative to the Area Median Income (AMI), which is based on the median for a family
household of four people. While not as immediately intuitive as categories based on absolute income, the
advantage of this approach is its sensitivity to household size, such that a large and small household with
comparable incomes may be in different categories reflecting the difference in the cost of living. For the
2010 base year, the median income in the City of Boston ($53,000) is approximately 60% of the regional
median income used by HUD ($91,800). The income breaks by household size are presented in the table
below. To project future year households by income, MAPC assumed that the distribution across income
levels would remain consistent for each combination of householder age and household type.
Household
Size
<=30% AMI
30-50% AMI
50-80% AMI
80-100% AMI
100-120%
AMI
120-150%
AMI
$19,300
$32,150
$45,100
$ 64,300
$ 77,160
$ 96,450
$22,050
$36,750
$51,550
$ 73,500
$ 88,200
$ 110,250
$24,800
$41,350
$58,000
$ 82,700
$ 99,240
$ 124,050
$27,550
$45,900
$64,400
$ 91,800
$ 110,160
$ 137,700
$29,800
$49,600
$69,600
$ 99,200
$ 119,040
$ 148,800
$32,000
$53,250
$74,750
$ 106,500
$ 127,800
$ 159,750
$34,200
$56,950
$79,900
$ 113,900
$ 136,680
$ 170,850
8 plus
$36,400
$60,600
$85,050
$ 121,200
$ 145,440
$ 181,800
In addition to meeting the growth in housing demand, additional units are needed to achieve and maintain
a healthy vacancy rate. How many additional units depends on what the target vacancy is. As shown on
the chart below, rental vacancy rates and numbers of for-rent units are at their lowest point since 1970.
MAPC has previously established target vacancy rates of 1.5% for ownership units and 7% for rental
units. These were used as the target vacancy rate assumptions. However, given that Boston has not
achieved a 7% vacancy rate in the rental markets since 1990, MAPC also developed housing demand
scenarios using a moderate rental vacancy rate of 6%.
Page 4
25,000
Vacant- Other
20,000
Vacant- Seasonal
15,000
1.8%
10,000
2.6%
Vacant- Rented or
Sold, not occupied
0.9%
2.2%
1.7%
7.5%
5,000
7.8%
6.0%
1.0%
5.2%
2.3%
0.4%
0.8%
4.6%
3.0%
4.2%
1.3%
3.6%
2.6%
2.2%
2014
2015
1970
1980
1990
2000
2010
2011
2012
2013
70,000
60,000
50,000
40,000
30,000
35,100
37,400
23,300
23,300
22,500
20,000
10,000
20,900
Scenario A
Scenario B
For-sale Units
Scenario C
Rental Units
Page 5
While these three scenarios illustrate the range of housing demand that may be associated with the BPDAs
new population projections, it is also necessary to adopt a single set of projections to use for planning
purposes. MAPC believes that the 2014 headship rates used for Scenario A reflect household formation
patterns during a housing cost crisis, and not actual preferences that should be used for planning purposes.
In other words, it is possible that the projected population of 2030 might fit in only 43,000 additional
housing units, but it would certainly be considered suboptimal for many residents, and might discourage
young workers from moving to the city. City housing policy should aim for production levels that enable
more choice in household formation. On the high end of the range, while the 7% vacancy rate in for-rent
units would be considered ideal, it may also be unrealistic given the current historically low vacancy rates
and the likelihood that higher vacancy rates could also slow down production. Given these considerations
about Scenarios A and B, MAPC recommends Scenario B (58,400 new units) as the appropriate
housing unit production goal to accompany the new BPDA population projections being used for
Imagine Boston. For the remainder of the report, this scenario is referred to as the Imagine Boston
Housing Projections Update.
The Imagine Boston Housing Projections Update anticipates that the number of households in the city will
grow to 307,800 in 2030, up from 252,700 in 2010. This results in a net growth of 55,200 householders
between 2010 and 2030, with about 60% of that growth occurring during the first forecast decade (see
chart below). To achieve and maintain the specified vacancy rates (1.5% for-sale units, 6% rental units)
the number of vacant units in the city will need to increase by 3,200 over the 2010 figure. Together the
demand for both occupied and vacant units equals 58,400 units.
320,000
307,800
300,000
285,200
280,000
252,700
260,000
239,600
240,000
261,500
228,000
220,000
200,000
1990
2000
2010
2015
2020
2030
The projected change in households by age of householder is shown in the chart below. Strong growth in
households headed by a person age 25 44 will account for 52% of the net change. Senior-headed
households are also projected to increase by about 28,500, as the citys large cohort of Baby Boomers
ages into its senior years. (This growth should not be interpreted to mean that large numbers of seniors are
projected to move into the city.)Meanwhile, MAPC projects that the number of householders under the age
of 25 is likely to decline as the Millennials age out of this group and more college students are housed in
dormitories. The number of householders age 45 64 is expected to grow by only 6,600.
Page 6
20,000
15,000
10,000
18,331
12,218
5,000
-
3,549
580
2,489
45-54
55-59
60-64
14,301
14,209
65-74
75P
-7,242
(5,000)
(10,000)
15-24
25-34
35-44
Age of Householder
The chart below depicts the net change in households by income level and age of householder, assuming
that income distribution by age and type of household remains as it is today. Under those conditions, about
half of the net change (equivalent to 30,500 households) would be in low-income households (earning less
than 80% of AMI.) Growth in senior-headed households accounts for more than two-thirds of that growth in
low-income households. For net change in households headed by residents under 65, the income distribution
is much more balanced: 10,300 low-income households and 12,200 high-income households (above 120%
AMI.)Of course, changing preferences and rising housing prices may result in faster growth of higher
income households and slower growth of lower income households, but these figures indicate the type of
housing that would be needed to maintain the citys current income distribution.
60,000
50,000
120-150% AMI
40,000
100-120% AMI
30,000
80-100% AMI
20,000
50-80% AMI
10,000
30-50% AMI
Less than 30% AMI
All Households
Householder <65
Householder 65 plus
Page 7
4+ people
15,000
2 - 3 people
10,000
One Person
5,000
(5,000)
Family
Non-Family
Page 8
for smaller housing units. Only 14% of the net household growth will be large families of four or more
people. While low income households comprise a large share of the net new households, most of that
growth is due to increases in the low-income senior population, highlighting the need for policies to support
seniors on fixed incomes in both rental and owner-occupied housing. The income distribution of workingage households is more balanced, with 35% low-income and 41% high income, suggesting a need for
production across the income spectrum.
Credits and Acknowledgements
This report was prepared by the MAPC Data Services Department, with analysis and modeling conducted
by Senior Research Analyst Meghna Hari. The report was authored by Data Services Director Tim
Reardon. We acknowledge and appreciate the collaboration and insight of the BPDA Research Division,
specifically Jonathan Lee, Christina Kim, Matthew Resseger, and Phillip Granberry.
Page 9