2017 City Demographic Estimates Continue to Show a Changing City

The 2017 Census estimates came out today for cities and counties. The estimates can be found here.

Highlights for the City of Columbus
-The non-Hispanic Asian population continues to skyrocket, up over 67% since 2010.
-Beyond that, all other racial groups saw population growth within the city since 2010.
-The foreign-born population has climbed above 105,000, and now represents 12% of the total population, the highest % level since 1890.
-Every age group has increased since 2010, but the older working-age population increased the most, as seen below:
19 and Under: +17,962
20-34: +22,627
35-64: +32,045
65+: +22,234

Check out all of the City, County and Metro Area demographic and population data on the Columbus Demographics page.

Census Tract Population Density 2015

The US Census recently released population data for census tracts. I figured midway through the decade would be a good point to update where these stand because they give greater insight in smaller-scale population changes. I looked at all the census tracts in Franklin County and came up with the following map series.

First, the population in 2015.

Next, the population density of tracts in 2010, as reference.

And now 2015.

On the surface, it’s difficult to see the changes, but put side by side, you can tell there have been a lot of increases across the county. To make this more visible, I made the following maps.

You can see that some of the strongest density increases occurred around Downtown and the Short North, New Albany, parts of the Campus area, and Dublin.

The map above gives a straightforward look at where the density increased and decreased. As you can see, the increases FAR outweighed the decreases. Most of the latter were scattered except across the Far South Side and parts of the Whitehall area.

Here were the top 20 most dense census tracts in 2015.
1. 1810: 29,508.2 South Campus/Victorian Village
2. 1121: 25,287.9 Main Campus
3. 13: 21,961.4 Campus/Indianola Terrace
4. 1110: 18168.6 North Campus/Tuttle Park
5. 10: 17386.3 Campus/SoHud
6. 12: 16,981.9 Campus/Iuka Ravine
7. 20: 13,030.5 Short North/Victorian Village
8. 17: 12,872.3 Weinland Park
9. 6: 12,153.6 Old North Columbus
10. 21: 10,853.5 Short North/High Street
11. 8163: 10,255.3 Lincoln Village/Southwest Columbus
12: 4810: 9,557.4 South Central Hilltop
13. 47: 9,492.7 North Central Hilltop
14. 6352: 9,434.0 Northwest Columbus/Henderson Road
15. 57: 9,257.4 Brewery District/South German Village
16. 5: 9,177.9 Old North Columbus
17. 6933: 9,090.9 Forest Park East
18. 16: 8,980.5 Weinland Park
19. 4620: 8,928.6 North Central Hilltop
20. 1820: 8743.3 Victorian Village

It’s obvious that the High Street corridor is the most dense of the city, racking up most of the top 20.

Now here are the 20 tracts with the largest density increases 2010-2015.
1. 1121: 4,375.9
2. 6: 2,178.5
3. 21: 1,934.9
4. 22: 1,478.1
5. 40: 1,107.7 South Downtown
6. 1820: 1,044.1
7. 20: 921.7
8. 38: 904.3 Old Towne East
9. 5: 861.2
10. 210: 833.9 Clintonville
11. 32: 751.1 Arena District West/West Victorian Village
12. 730: 736.9
13. 7551: 656.0 Somerset/South Easton
14. 7951: 610.4 West Columbus
15. 6372: 574.6 Hayden Falls/Sawmill Road
16. 7209: 514 New Albany
17. 7395: 497.6 Blacklick/East Broad
18. 10: 492.8
19. 8230: 449.3 Westland
20. 710: 447.3 West-Central Linden

And finally, the top 20 largest declines 2010-2015.
1. 13: -2,964.3
2. 12: -1,625.1
3. 42: -1,620.8 Scioto Peninsula/East Franklinton
4. 920: -902.2 Northeast Linden
5. 17: -775.4
6. 50: -554.4 Franklinton
7. 61: -485.7 South High Street
8. 59: -441.9 Near South Side/Deshler Park
9. 4620: -380.4
10. 720: -380.2
11. 4610: -335.4
12. 820: -305.4 North Linden
13. 7721: -305.2 North Linden
14. 45: -258.1 North Hilltop
15. 60: -253.2 Vassor Village
16. 810: North Central Linden
17. 7532: -240.3 Morse Road/Easton
18. 2520: -240.1 Near East Side/King-Lincoln
19. 47: -206.6
20. 9333: -194.9 Linwood

So there you have it.

Housing Trends of Columbus

***Originally Posted May 23, 2014, updated with 2014 data 9/18/2015 and again on 5/29/2016 with 2015 data***

I posted a graph recently showing housing permits for Franklin County to show how construction was trending. Today, I found more long-term data for both the city and county that continue to show some interesting trends.

First, let’s look at just the city of Columbus.

The chart above goes back through the mid-1990s. The first thing to notice is the housing boom from 1999-2002. Both single-family and multi-family construction was booming. The very good economic conditions, or seemingly good ones, during the 1999-2000 period is probably most responsible for this. What’s most interesting is that the boom seemed to last through at least part of the mild recession experienced in 2001-2002. After that, housing of both types started to decline through the late 2000s. This shows that construction in the city began to decline as early as 2002-2003, before the peak of the general housing boom in the mid-2000s.

Another interesting fact is at the end of the period. Multi-family units have recovered and are back in boom territory. This boom, however, is much different than the one that occurred more than a decade ago, as shown by the below chart.

During the 1999-2002 housing boom, multi-family housing averaged 59.3% of all the units constructed. In the current boom, which began in 2012, multi-family housing has averaged 81.4% of all the units constructed. The average difference between the types 1999-2002 was just 18.6 points. In the current boom, the difference is almost 63 points! In that regard, there really is no comparison between the housing boom a decade ago and the current one. Multi-family construction is in MUCH higher relative demand now than it was at any time in the last 20 years, including during the last housing boom.

But what does this tell us about where the housing is actually being constructed? Well, for that, we have to look at the entirety of Franklin County. Is the county also seeing a similar multi-family boom, or has single-family construction recovered there more than in the city?

This chart, in some aspects, is the opposite of the one for the city. While in the city, multi-family units consistently outnumbered single-family, the opposite is true for the county as a whole. This is likely because the county takes into account all the suburban areas, most of which are dominated by single-family housing. In only a few instances did multi-family housing units outnumber single-family before 2010. After 2010, it’s clear that the multi-family boom is hitting the rest of the county and not just Columbus itself. This may actually represent an even greater shift in housing construction. While it appeared that single-family construction was gradually rising since 2011, it once again fell off some in 2015 while multi-family went up. It appears that the new reality is, at least for now, holding steady.

Here’s the % of total chart for the county.

So it’s also clear that the county is seeing most of its construction in recent years be multi-family units.

Economic Segregation in Columbus

Luckily, I saved all of my maps that I did for this report, so not all was lost. Instead of making it a 2-part post, I’m just reposting it all in one this time.

In any case, economic segregation is basically where people living in the same city are segregated in terms of financial characteristics, such as housing prices or income. This is considered negative as the more economically segregated an area is, the harder it is for people, especially in lower income brackets, to move up financially. My report focuses on household income within census tracts in Franklin County and where those household incomes are changing the most.

First of all, let’s look at the household income levels around the county, both in 2000 and 2014.

In 2000, the median household income for the county was highest in the Upper Arlington and Grandview, Dublin, Bexley, Hilliard and the New Albany area. Downtown and adjacent areas had the lowest, as well as the general urban core and East Side.

By 2014, household income remained the highest in the same areas it was in 2000, but there were major improvements in many parts of the urban core, especially around Downtown, the Near East Side, Near South, Clintonville and the Short North. To illustrate this change better, take a look at the next map.

Unfortunately, because not all of 2014’s census tracts existed in 2000, I don’t have data for the entire county for comparison. But the trend is very clear. The areas that saw the biggest improvements in median household incomes were in the dead center of the county- Downtown, Near South and East Sides, as well as the Short North and Grandview. Only parts of Hilliard, Clintonville and Worthington really saw anything remotely as close. This indicates, at least to me, that the beating heart of revitalization and growth in the county is along the High Street corridor.

So now that we’ve established what the incomes look like across the county, let’s break it down further into income level brackets. This will help determine where economic segregation is a problem and where it isn’t.

The lowest household income I looked at was Below $25K a year. In 2000, this income level was most heavily concentrated in the Downtown area and adjacent neighborhoods. The Near East Side, as well as Linden down through the east side of I-71 had the county’s highest % of households that earned this level of income. Hilltop and the West Broad Corridor were also fairly high.

By 2014, the lowest household income level looked largely the same. However, there were also some noticeable difference. Downtown, the Near East Side, the Near South Side and parts of the North High Corridor saw obvious declines in this population, while it seemed to spread further east outside of 270 into suburban areas.

In the map above, we can see how Below $25K household incomes had changed in the tracts between 2000 and 2014 by % point change. Ironically, the urban core, especially along High and Broad streets saw the most consistent declines in this population while areas around and outside of 270 saw the most consistent increases. The good news is that more tracts saw declines than increases, but the map does indicate that poverty is perhaps moving further out from the core.

Next up is the household income level change that would be considered closest to middle class- $50K-$99K.

The urban core areas clearly saw the most consistent increases in middle class household income levels, while the outer suburbs almost universally declined in this metric. One explanation for this is that the lowest incomes in the core moved up into the middle class, while in the suburbs, middle class incomes moved into the upper class incomes. That would explain both the rise in the core, but the decline in the suburbs. But to prove if this is true or not, we have to look at the highest income levels- those of $100K and above.

In 2000 the highest incomes were almost entirely outside of 270 except for Bexley and the Northwest Side communities like Dublin and Upper Arlington. It is likely that the New Albany area also had high incomes, but again, those tracts didn’t exist in 2000, so it is difficult to give that information.

By 2014, while the Northern areas of Franklin County continued to have the highest incomes in general, gains were made in many parts of the county, including several within the urban core area.

Between 2000 and 2014, there was almost universal growth of $100K+ incomes in Franklin County, with only small areas seeing declines. The Northwest communities, as well as areas in and around Downtown seemed to do the best.

Okay, so incomes levels are clearly improving in most of the county, but especially in urban core areas. But what is the difference between the highest and lowest incomes within each census tract? To find out, I took the % of households in each tract earning less than $25K a year vs. the % of households earning $100K or more. The % point difference between these two groups is a good indication of how much economic segregation exists. The closer this number is to 0, the more economically integrated a tract is. Negative numbers indicate that Below $25K household incomes outweigh those making $100K or more, while positive numbers are the reverse.

The 2000 map shows that Below $25K household incomes dominate inside I-270, particularly around Downtown and the East Side. Many tracts contain at least 40 % points more $25K incomes than $100K incomes. This shows that poverty was deeply concentrated around the center of the county. Suburban areas were more dominated by the reverse, where middle and upper class households were concentrated.

In 2014, the severely concentrated levels of the lowest incomes have eased in most locations. There are fewer tracts of 40+ point differences, especially around Downtown and the general High Street Corridor. Only the Campus area, for obvious reasons, and parts of Linden, largely remain unchanged.

So what does all this ultimately mean about economic segregation in Frankly County? To get a simplified sense of that picture, considering the final set of maps.

In the coloring, the blue tracts are tracts that have income point differences that are between -15 and +15. These are the tracts that are most economically integrated. Green tracts are those with differences of +/- 15 to 29 points, while orange represent those with +/- 30 points or more. Orange tracts are the most economically segregated. In 2000, most of the orange tracts were within I-270. In fact, they very closely represent the most urban part of Columbus- the 1950 city boundary. They are amazingly similar. Meanwhile, almost all the outer suburbs in 2000 were well integrated.

Fast forward to 2014 and the picture becomes significantly more convoluted. Being in the urban core vs. the suburbs does not automatically guarantee economic integration. Many suburbs are now as severely segregated as some of the urban core is, while parts of the urban core are as integrated as some suburbs.

Overall, it appears that Franklin County has improved its economic integration in the last decade or so, but there is still more than can be done. Economic incentives for providing more mixed-income housing and bringing more jobs to urban areas would likely help achieve a more integrated city and county.

Cool Link of the Day: Mapping Commuting Patterns

Curious to know how people get to work in every county in the United States? Use the following link to find out.

The map was constructed using 2013 data, so it’s fairly recent. As for Franklin County? Here’s the breakdown:

Drive Alone: 82%
Carpool: 8%
Public Transit: 2%
Walk: 2%
Bicycle: 1%
Taxi or Other: 1%
Work from Home: 4%

The numbers are overwhelmingly auto-centric, as they are nearly everywhere, but what the numbers don’t show are any trends.