Recent history of Michigan housing trends

An overall look at the total number of housing units in Michigan shows there has been an increase in the number built from 1940 to 2010. However, in the remainder of this post information will show that housing construction decreased significantly beginning in 2009.

After a dip in the early 1990’s in the number of private housing units authorized, an indicator of the number of privately owned housing starts, Michigan began a slow, bumpy increase. Then, from 2004 to 2009 there was a steady and rapid decline in the number of building permits authorized. There was a drop of 45,000 building permits authorized from peak to bottom. In 2010 there was a slight increase, but permits still did not reach the 10,000 mark through 2011.

This chart shows the number of privately owned one unit housing starts from 1990 to 2011.  From 1990 to 2004 the number of privately owned housing starts showed growth. However, beginning in 2005 those numbers began to drop off and by 2009 that number had reached its all time low in the 21 years recorded at 6,441. The high during this time period was 44,186 in 2004.

The above chart shows the age of housing from data provided by both the 2000 and 2010 censuses. When comparing the age of housing from the 2000 and the 2010 data it can be seen that there was more new homes built from 1991 to 2000 than from 2001 to 2010. The data also shows that in 2010 the number of homes 50 years and older was much higher, by 576,531, in the year 2010 than 2000. This has huge implications for housing maintenance costs in the state at a time of declining incomes.

Lead poisoning and the City of Detroit

The United States Centers for Disease Control and Prevention (CDC) committed to the goal of eliminating blood lead levels of greater than 10 micrograms of lead per deciliter of blood (10 µg/dl) by 2020 and recommends public health actions at levels above 5µg/dl (see the CDC Lead Home Page). While lead poisoning has few physical symptoms, it can result in learning and behavioral disabilities. Children between the ages of six months and 3 years are at the greatest risk because they have a high degree of exposure to lead in household dust due to normal hand-to-mouth activity. Their developing nervous system is at heightened vulnerability to lead toxicity.

This chart shows the number of children (under 6 years old) newly identified with blood lead levels at or above 10 µg/dl over the past 14 years. This chart demonstrates a mostly steady decline in lead poisoning cases over the past 14 years. The small spike in 2001 and 2002 can be attributed to a new Medicaid policy enacted that required testing of blood lead levels among children considered to be at a high risk for lead poisoning. In 1998 9.7 percent (4,846 children) of children under the age of 6 in Detroit had blood lead levels at or above 10 µg/dl; in 2001 that decreased to 5.5 percent (3,236). High risks of lead poisoning can be typically be attributed to areas with lower income housing and homes that were built before 1978. Medicaid considers all children living in the City of Detroit to be at high risk for lead poisoning.

This map shows the locations of lead poisoning cases for children under 6 that were identified in Detroit it 2011. The green dots represent residences of children with blood lead levels between 10 to 19 µg/dL and the red dots are more severe  lead poisoning cases (20 to 99 µg/dL). Lead poisoning cases were recorded throughout most of Detroit, though there was a concentration in the center of the city (older housing) and on its East side.

The above map shows the number of lead poisoning incidences per address/home in the City of Detroit. The yellow dots show that there are 4,610 homes in the city with two occurrences per home; these are the most frequent. However, the blue dots, which show  three to five or six to nine occurrences per home, cover more of the map because of the higher number of people affected. There was one home in the southwest portion of the city that had 17 lead poisoning cases, according to the map.

The above chart shows the number of Michigan children under the age of 6 whose blood lead levels were tested over the past 14 years. The steady increase in the number of children tested can be at least partially attributed to Medicaid requirements.  The total number of children tested increased from 73,643 in 1998 to 151,867 in 2011. Most of this increase was due to children receiving Medicaid: in 1998, approximately 57 percent of the children tested were receiving Medicaid; in 2011, 76 percent were.

This chart shows the percentage of Michigan children under 6, of those who were tested, who had elevated blood lead levels over the past 14 years. As depicted in the orange line, nearly 10 percent of children tested had blood lead levels of at least 10 µg/dL in 1998. In 2011, less than 1 percent of children tested had levels of at least 10 µg/dL. The middle line (brown) represents more severe cases (at least 15 µg/dL) and the red line represents the most severe (at least 20 µg/dL). The decline in lead poisoning cases among all three categories can be attributed, directly and indirectly, to increases in testing among high risk children, increases in vacancy among the oldest, most dangerous and hazardous housing stock, and targeted lead poisoning prevention and case management.

Taking stock of Detroit neighborhoods, July 2012

Detroit’s Address Vacancy Rate Increases

The latest analyses of the Federal Reserve’s Open Market Committee (the Fed) indicates that nationally the residential real estate sector remains weak for the most part. However, the Fed points out that some local real estate markets are doing fairly well. Nevertheless, the housing market as a whole continues to face three major challenges:

1)a continuing large number of foreclosures and other distressed properties;
2)uncertainty about the future of home prices; and,
3)tight underwriting standards for mortgage loans.

One measure that can be used to judge trends in local real estate markets is the share of buildings that are vacant at a particular moment in time. The newest source of vacancy data comes from the US Postal Services (USPS) via the Department of Housing and Urban Development (HUD).  HUD recently entered into an agreement with the USPS to aggregate and publicly release Postal Service data on vacant addresses on a quarterly basis.  The  USPS data covers the entire universe of

Using the 25 HUD /USPS extracts since December 2005,  a time-series database consisting of over 1 million records that describe the address vacancy conditions in every census tract in the United States has been created.

There are a number of basic measures for this data set including:

•Total number of residential addresses;
•Total number of residential, business and other addresses that are vacant;
•Vacant , as defined by the USPS, means the resident(s) have not collected their mail for 90 days or longer.
•Total number of addresses that are “no-stat” (i.e., temporary vacancies).
•The “no-stat” category is separate from vacant; it includes: a) addresses under construction not yet occupied; b) rural addresses  vacant for 90 days or longer; c) urban addresses identified by a carrier as not likely to be active for some time (e.g., if a building being  demolished is to be replaced by another building, the address is preserved and considered “no-stat”).

The following  figures and maps are based on the USPS/HUD data.

In this chart, quarterly vacancy data begins in December 2005, which is the row of data labeled as “1205” under the column header “Quarter”.  In March 2008, the USPS began to differentiate residential addresses from businesses and other types of addresses, which begin with the fifth column labeled “Residential”.

For March 2012, the overall vacancy rate increased from 20.93 to 21.19 – as 833 addresses became vacant.  The vacancy rate for each property type (residential, business and other) vacancy increased slightly from December 2011. For all quarters in which the USPS differentiated between types of addresses, the vacancy rate for businesses has been higher than the vacancy rate for residences.

The overall address vacancy rate in the City of Detroit has increased from just over 10 percent during 2005 to just over 21 percent for 2012.

In March 2010, the USPS made changes to its counting procedures to improve the accuracy of the vacant indicator. While these new methods should improve data quality over time, caution should be used in measuring change over time. For that reason, it can be generally observed that vacancy in Detroit is increasing, but it may not be as straightforward to make year-to-year comparisons.   For more information, see the following http://www.huduser.org/portal/datasets/usps.html.

The map above of census tracts in the City of Detroit illustrates spatial patterns in overall address vacancy rates for March 2012 .  There are clusters of neighborhoods (shaded purple) with vacancy rates exceeding 30 percent, including the Brightmoor area on Detroit’s west side, the State Fair-Nolan area, the Tireman area, and numerous neighborhoods on Detroit’s east side; a few have vacancy rates exceeding 50 percent.

This map of the City of Detroit shows the change in address vacancy rates by census tract from March 2011 to March 2012. The red areas indicate areas showing increases in address vacancy by more than 1 percent during the past year. Green areas showed declines in vacancy rates (improvements) for this same period.

Areas shaded in gray showed little or no change in the past year (i.e., +/- 1 percent).  As illustrated, several areas on Detroit’s west side showed signs of improvement, including the Grandmont/Rosedale and 7 Mile/Greenfield neighborhoods.  Other improvements included a big decline in address vacancy in Midtown and some modest improvements in areas along East Jefferson Avenue, on Detroit lower east side.

Detroit asthma rates compared to state, national levels

The following tables and maps show the prevalence of asthma at both the state level for Michigan and for the City of Detroit. While Detroit is the main focus of this post, the charts that show asthma indicators and prevalence rates also provide insight on asthma data at the state and national levels.

The table above, Table 1 ,shows various asthma indicators for a study of children (under 18 years of age) enrolled in Michigan Medicaid programs in 2010.  The City of Detroit had higher numbers for every indicator listed above than those at the state level.

Data definitions are presented below:

Persistent Asthma:  The proportion of the study population who meet the definition of persistent asthma in 2010.

Emergency Department Rate:  The number of emergency department visits with a primary diagnosis of asthma per 10,000 children continuously enrolled in the Michigan Medicaid population in 2010. These data represent the number of visits for asthma, not the number of persons with a particular type of event.

Asthma Hospitalization Rate:  The number of in-patient hospitalizations with a primary discharge diagnosis of asthma per 10,000 children continuously enrolled in the Michigan Medicaid population in 2010. These data represent the number of hospitalizations for asthma, not the number of persons hospitalized.

Asthma Utilization Indicators among children with persistent asthma in Michigan Medicaid programs: Four indicators are presented to capture an element of asthma care or management among children who have asthma in this population.

  • Percent of Children with Persistent Asthma who had at least one Emergency Department Visit due to Asthma.
  • Percent of Children with Persistent Asthma who had two or more Emergency Department Visits due to Asthma.
  • Percent of Reliance on Emergency Department for outpatient care. (ED reliance is an indicator of outpatient utilization for asthma within an emergency department as opposed to in an outpatient setting)
  • Percent of children with persistent asthma who had at least one hospitalization due to asthma during 2010.

The above table, Table 2, shows the asthma prevalence rate at the country and state level for children in four different age groups in 2010. Except for the 10-14 years old age group, asthma prevalence rates in Michigan were higher than the national rates.

Data definitions are presented below:

Prevalence Rate:  The estimated proportion of the population with asthma in 2010.

Prevalence Number:  The estimated number of individuals in the population with asthma in 2010.

The above map shows persistent asthma prevalence rates for children (under 18 years of age) enrolled in Medicaid in 2010 for each zip code in the City of Detroit. The asthma prevalence rate for children in Medicaid in the City of Detroit in 2010 ranged from 4.6 percent to 7.9 percent. Five zip codes recorded the highest rates (6.71%-7.9%) among all 27 zip codes.  These five zip codes are: 48201, 48207, 48212, 48213, and 48216. As depicted in Table 1, the persistent asthma prevalence rate for children (under 18 years of age) enrolled in Medicaid in the entire City of Detroit in 2010 was 6 percent.

This map shows a combination of three indices of asthma burden in Detroit by zip code. The lower x-axis (1) shows persistent asthma prevalence rates for children (under 18 years of age) enrolled in Medicaid in Detroit in 2010. The y-axis (2) shows the number of in-patient hospitalizations with a primary diagnosis of asthma per 10,000 children (under 18 years of age) enrolled in Medicaid in Detroit in 2010. The upper x-axis (3) shows the number of emergency department visits with a primary diagnosis of asthma per 10,000 children (under 18 years of age) enrolled in Medicaid in Detroit in 2010. The map in the upper right corner shows the three zip codes with the highest rates of asthma prevalence, hospitalizations, and emergency department visits in Detroit in 2010. These three zip codes are  48201,48207, 48213; they are also three of the five listed as having the highest asthma prevalence rate in the City of Detroit in 2010.


Detroit’s population density

From 1950 to 2010, Census tract data indicates there has been a steep population decrease in the City of Detroit. The population loss began in the center of Detroit, which in the first map can be seen as the area with the highest population density. However, as time progressed the city’s population density continued to diminish. By 1980, there were such significant population losses in sections of the city that the population density was recorded as zero in one area, which is east of southbound I-75 and south of I-94. By 2010 this area did not show population growth, and the city’s overall population continued to decrease. 

While the maps above provide a strong visual indication of  Detroit’s population loss, the line graph below shows the steady decline in  population density. According to the chart, there were 13,330 people per square mile in Detroit in 1950, and by 2010 that number had decreased to 5,144.