As a lifelong resident of coastal/metro California, this definitely surprised me. When I think of the most affluent neighborhoods, it's places like Pacific Heights, Hillsborough, Atherton, Piedmont, Hancock Park, Bel Aire, Beverly Hills . . . all largely built out at least 100 years ago. But yes, I guess those are really the exception nationally.
Yes! California is a really interesting case and unique because all the state's constraints on new housing have pushed a lot of investment into the maintenance and upkeep of older stock. In some ways, it shows how, when you can't build new, there's a lot more reinvestment in old. But securing neighborhood well-being through housing scarcity isn't a solution -- you end up with the affordability crisis plaguing the state.
I do not take a study seriously which divides a population into quintiles and then labels the bottom one “distressed”. It is just question begging in the form of empirical number crunching.
People without college, lower than average incomes, and Hispanic heritage need to live somewhere too. But then you label them as distressed despite the fact that they proudly live in well maintained homes worth well over a million dollars (SoCal).
Absolutely worse than worthless. Studies like this effectively destroy knowledge more than they create it.
It's fair to critique the labeling but we'll have to agree to disagree on the value of measuring how relative economic and social well-being varies from one neighborhood to the next. From health outcomes to school quality and job opportunities, almost any dimension of life we care about will vary from one place to the next and be highly correlated with the indicators we wrap into the DCI. The zip codes we label as distressed lost jobs and saw more business close than open over the past 5 years, when the U.S. economy was booming. I personally think it's vitally important to know who our economy is failing and where.
Sorry for the long answer, but I just ran this index methodology through paid AI and it mentions exactly what I said, but with more thorough arguments.
This is an advocacy program disguised as an empirical measurement system.
Here is Chao GPT;
The underlying data are useful, but the categories—prosperous, comfortable, mid-tier, at risk, distressed—turn a relative ranking into an apparently objective description of community conditions.
The central problem: these are forced quintiles
The DCI ranks communities nationally on seven indicators, averages their percentile ranks, and then divides them into five equal groups:
Thus, exactly 20% of the ranked communities must be called distressed and another 20% at risk, regardless of how prosperous the country becomes or how small the actual differences between communities are. Likewise, exactly 20% must be called prosperous even if economic conditions were broadly terrible.
Those are not natural or empirically discovered thresholds. They are simply labels attached to ranks.
A more accurate naming system would be:
Highest-performing quintile
Above-average quintile
Middle quintile
Below-average quintile
Lowest-performing quintile
That sounds less dramatic because it accurately describes what the calculation does.
“At risk” is especially misleading
“At risk” sounds predictive: at risk of decline, poverty, fiscal collapse, unemployment, or some identifiable adverse event.
But membership in that category does not necessarily mean the community is deteriorating or faces an elevated probability of future distress. It means it ranked between approximately the 60th and 80th percentiles of the index’s particular conception of distress.
A stable, pleasant, middle-income community can therefore be designated “at risk” simply because 60% of the comparison group scores better.
The index does include two dynamic variables—five-year employment growth and business-establishment growth—but five of its seven indicators describe current conditions. The composite has not, merely by including two growth measures, been demonstrated to be a risk-prediction model.
“Distressed” also overstates the finding
Some places in the bottom quintile genuinely are deeply distressed: concentrated poverty, low labor-force participation, abandoned housing, and disappearing businesses. In those cases the word is apt.
But the methodology does not first establish an absolute threshold for distress. A community is “distressed” because it is in the bottom fifth relative to other communities.
That conflates two different propositions:
This community performs poorly relative to the national distribution.
This community is experiencing serious economic distress.
The first is established by the index. The second may be true, but it requires either an absolute standard or additional evidence.
“Prosperous” and “comfortable” contain the same error
A ZIP code in the top quintile is not necessarily prosperous in any ordinary sense. It may have modest incomes but low poverty, high employment, few vacant houses, and decent recent growth. “Relatively high-scoring” would be defensible; “prosperous” is an interpretation.
“Comfortable” is even stranger. Comfort is a broad human-welfare condition involving:
disposable income and cost of living,
housing costs and quality,
health and safety,
household wealth and debt,
leisure and job security,
amenities and environmental quality.
The DCI does not measure most of those things. Calling a place “comfortable” anthropomorphizes an economic rank.
The indicators reflect a particular theory of a healthy community
The seven variables are:
adults without a high-school diploma,
housing vacancy,
prime-age adults not working,
poverty,
local median income relative to the surrounding area,
employment change,
business-establishment change.
These are reasonable indicators of economic weakness. But they privilege a specific model: a successful community is educated, employed, occupied, relatively high-income, and growing in jobs and establishments.
That creates several distortions.
Growth is treated as well-being
A mature, stable community with little population or business growth can be marked down relative to a booming exurban area. Yet rapid growth may bring congestion, housing shortages, infrastructure stress and declining affordability. Conversely, stability is not necessarily distress.
Relative income can produce odd results
The current methodology defines median income relative to the metropolitan or state median, which helps control for regional differences. But it also means that a community’s score depends partly on the prosperity of its neighbors. A given household income can look strong in one region and weak in another.
That is defensible for measuring relative local disadvantage, but less defensible if the resulting category is presented as an absolute statement about well-being.
Equal weighting is arbitrary
Each of the seven indicators receives equal weight. That implies, for example, that business-establishment growth has the same conceptual importance as poverty or prime-age nonemployment. EIG may have practical reasons for simplicity, but equal weighting is not a neutral empirical discovery.
It measures places, not necessarily residents’ welfare
A thriving business district can coexist with expensive housing and financially strained residents. A retirement community may have low employment participation for benign reasons. A college town can look peculiar on poverty and income statistics—which is one reason EIG excludes ZIP codes where students constitute at least half the population.
The index is therefore a local economic-structure index, not a comprehensive measure of how well residents are living.
Is the naming politically biased?
I would call it advocacy-biased rather than clearly partisan-biased.
EIG works on policies intended to stimulate investment and entrepreneurship in economically lagging places. Dramatic categories make the map legible, memorable, and policy-relevant. “Forty percent of communities are below the national median” has little emotional force. “Forty percent are distressed or at risk” sounds like a problem demanding intervention.
The rhetorical direction is therefore predictable:
relative disadvantage becomes “distress,”
below-average performance becomes “risk,”
above-average performance becomes “comfort” or “prosperity.”
The labels embed a narrative of crisis and remedy into what is fundamentally a percentile-ranking exercise.
My overall verdict
The DCI is a useful comparative dashboard wrapped in misleading categorical language.
Its strongest defensible claim is:
This index shows how communities rank relative to one another on seven selected measures of economic disadvantage and local economic dynamism.
Its presentation instead encourages the stronger interpretation:
This index identifies which communities objectively are prosperous, comfortable, endangered, or distressed.
That stronger interpretation is not warranted by the methodology.
The greatest flaw is not that the measurements are obviously manipulated. It is that ordinal rankings have been converted into qualitative diagnoses without absolute thresholds, validation of the category boundaries, or sufficient acknowledgment that the labels are conventional rather than discovered.
It's an empirical observation about a clear and pervasive relationship across the country's map. I'm sorry it didn't resonate with you and appreciate the book recommendation, but I don't see what's so contrary to the Strong Towns thesis. I am not endorsing the Sun Belt sprawl model, I'm observing that we've broadly failed to find a model of urban development that sustains high levels of well-being across other types of communities too, especially as they grow older. In a country as diverse and dynamic as ours, there are obviously exceptions -- and hopefully we can learn from them to make a country of stronger places. And yes, there are compositional effects to unpack. The piece acknowledges them but presses on because the regularity of the relationship itself is important and noteworthy -- and necessary to make before disentangling the forces underlying it.
As a lifelong resident of coastal/metro California, this definitely surprised me. When I think of the most affluent neighborhoods, it's places like Pacific Heights, Hillsborough, Atherton, Piedmont, Hancock Park, Bel Aire, Beverly Hills . . . all largely built out at least 100 years ago. But yes, I guess those are really the exception nationally.
Yes! California is a really interesting case and unique because all the state's constraints on new housing have pushed a lot of investment into the maintenance and upkeep of older stock. In some ways, it shows how, when you can't build new, there's a lot more reinvestment in old. But securing neighborhood well-being through housing scarcity isn't a solution -- you end up with the affordability crisis plaguing the state.
I do not take a study seriously which divides a population into quintiles and then labels the bottom one “distressed”. It is just question begging in the form of empirical number crunching.
People without college, lower than average incomes, and Hispanic heritage need to live somewhere too. But then you label them as distressed despite the fact that they proudly live in well maintained homes worth well over a million dollars (SoCal).
Absolutely worse than worthless. Studies like this effectively destroy knowledge more than they create it.
It's fair to critique the labeling but we'll have to agree to disagree on the value of measuring how relative economic and social well-being varies from one neighborhood to the next. From health outcomes to school quality and job opportunities, almost any dimension of life we care about will vary from one place to the next and be highly correlated with the indicators we wrap into the DCI. The zip codes we label as distressed lost jobs and saw more business close than open over the past 5 years, when the U.S. economy was booming. I personally think it's vitally important to know who our economy is failing and where.
Sorry for the long answer, but I just ran this index methodology through paid AI and it mentions exactly what I said, but with more thorough arguments.
This is an advocacy program disguised as an empirical measurement system.
Here is Chao GPT;
The underlying data are useful, but the categories—prosperous, comfortable, mid-tier, at risk, distressed—turn a relative ranking into an apparently objective description of community conditions.
The central problem: these are forced quintiles
The DCI ranks communities nationally on seven indicators, averages their percentile ranks, and then divides them into five equal groups:
Thus, exactly 20% of the ranked communities must be called distressed and another 20% at risk, regardless of how prosperous the country becomes or how small the actual differences between communities are. Likewise, exactly 20% must be called prosperous even if economic conditions were broadly terrible.
Those are not natural or empirically discovered thresholds. They are simply labels attached to ranks.
A more accurate naming system would be:
Highest-performing quintile
Above-average quintile
Middle quintile
Below-average quintile
Lowest-performing quintile
That sounds less dramatic because it accurately describes what the calculation does.
“At risk” is especially misleading
“At risk” sounds predictive: at risk of decline, poverty, fiscal collapse, unemployment, or some identifiable adverse event.
But membership in that category does not necessarily mean the community is deteriorating or faces an elevated probability of future distress. It means it ranked between approximately the 60th and 80th percentiles of the index’s particular conception of distress.
A stable, pleasant, middle-income community can therefore be designated “at risk” simply because 60% of the comparison group scores better.
The index does include two dynamic variables—five-year employment growth and business-establishment growth—but five of its seven indicators describe current conditions. The composite has not, merely by including two growth measures, been demonstrated to be a risk-prediction model.
“Distressed” also overstates the finding
Some places in the bottom quintile genuinely are deeply distressed: concentrated poverty, low labor-force participation, abandoned housing, and disappearing businesses. In those cases the word is apt.
But the methodology does not first establish an absolute threshold for distress. A community is “distressed” because it is in the bottom fifth relative to other communities.
That conflates two different propositions:
This community performs poorly relative to the national distribution.
This community is experiencing serious economic distress.
The first is established by the index. The second may be true, but it requires either an absolute standard or additional evidence.
“Prosperous” and “comfortable” contain the same error
A ZIP code in the top quintile is not necessarily prosperous in any ordinary sense. It may have modest incomes but low poverty, high employment, few vacant houses, and decent recent growth. “Relatively high-scoring” would be defensible; “prosperous” is an interpretation.
“Comfortable” is even stranger. Comfort is a broad human-welfare condition involving:
disposable income and cost of living,
housing costs and quality,
health and safety,
household wealth and debt,
leisure and job security,
amenities and environmental quality.
The DCI does not measure most of those things. Calling a place “comfortable” anthropomorphizes an economic rank.
The indicators reflect a particular theory of a healthy community
The seven variables are:
adults without a high-school diploma,
housing vacancy,
prime-age adults not working,
poverty,
local median income relative to the surrounding area,
employment change,
business-establishment change.
These are reasonable indicators of economic weakness. But they privilege a specific model: a successful community is educated, employed, occupied, relatively high-income, and growing in jobs and establishments.
That creates several distortions.
Growth is treated as well-being
A mature, stable community with little population or business growth can be marked down relative to a booming exurban area. Yet rapid growth may bring congestion, housing shortages, infrastructure stress and declining affordability. Conversely, stability is not necessarily distress.
Relative income can produce odd results
The current methodology defines median income relative to the metropolitan or state median, which helps control for regional differences. But it also means that a community’s score depends partly on the prosperity of its neighbors. A given household income can look strong in one region and weak in another.
That is defensible for measuring relative local disadvantage, but less defensible if the resulting category is presented as an absolute statement about well-being.
Equal weighting is arbitrary
Each of the seven indicators receives equal weight. That implies, for example, that business-establishment growth has the same conceptual importance as poverty or prime-age nonemployment. EIG may have practical reasons for simplicity, but equal weighting is not a neutral empirical discovery.
It measures places, not necessarily residents’ welfare
A thriving business district can coexist with expensive housing and financially strained residents. A retirement community may have low employment participation for benign reasons. A college town can look peculiar on poverty and income statistics—which is one reason EIG excludes ZIP codes where students constitute at least half the population.
The index is therefore a local economic-structure index, not a comprehensive measure of how well residents are living.
Is the naming politically biased?
I would call it advocacy-biased rather than clearly partisan-biased.
EIG works on policies intended to stimulate investment and entrepreneurship in economically lagging places. Dramatic categories make the map legible, memorable, and policy-relevant. “Forty percent of communities are below the national median” has little emotional force. “Forty percent are distressed or at risk” sounds like a problem demanding intervention.
The rhetorical direction is therefore predictable:
relative disadvantage becomes “distress,”
below-average performance becomes “risk,”
above-average performance becomes “comfort” or “prosperity.”
The labels embed a narrative of crisis and remedy into what is fundamentally a percentile-ranking exercise.
My overall verdict
The DCI is a useful comparative dashboard wrapped in misleading categorical language.
Its strongest defensible claim is:
This index shows how communities rank relative to one another on seven selected measures of economic disadvantage and local economic dynamism.
Its presentation instead encourages the stronger interpretation:
This index identifies which communities objectively are prosperous, comfortable, endangered, or distressed.
That stronger interpretation is not warranted by the methodology.
The greatest flaw is not that the measurements are obviously manipulated. It is that ordinal rankings have been converted into qualitative diagnoses without absolute thresholds, validation of the category boundaries, or sufficient acknowledgment that the labels are conventional rather than discovered.
Yeesh.
So many compositional effects to untangle. This is basically a worthless study.
Maybe go read Strong Towns and get back to us.
It's an empirical observation about a clear and pervasive relationship across the country's map. I'm sorry it didn't resonate with you and appreciate the book recommendation, but I don't see what's so contrary to the Strong Towns thesis. I am not endorsing the Sun Belt sprawl model, I'm observing that we've broadly failed to find a model of urban development that sustains high levels of well-being across other types of communities too, especially as they grow older. In a country as diverse and dynamic as ours, there are obviously exceptions -- and hopefully we can learn from them to make a country of stronger places. And yes, there are compositional effects to unpack. The piece acknowledges them but presses on because the regularity of the relationship itself is important and noteworthy -- and necessary to make before disentangling the forces underlying it.