There’s never enough parking.
We took away the curb and devoted it to on-street parking spaces which cost a pittance, if anything. We paved over vast swathes of green spaces, filled in wetlands, covered over streams. We bulldozed historic buildings and replaced them with endless surface lots and parking decks throughout our urban cores. We tore down houses, displaced vulnerable families, and closed small businesses to feed this insatiable demand for more parking.
But there can never be too much of a good thing.
Except when there is. Or when the thing isn’t actually some unvarnished good. And certainly when both of these things are true, as they are with parking.
Quantifying parking in American cities
There has been a remarkable amount of research done in the past few years on the scale and scope of the United States’ unchecked addiction to providing storage units for cars.
Back in 2003, Hashem Akbari and colleagues from the Lawrence Berkeley National Laboratory analyzed the Sacramento area using aerial photographs and concluded that roads, parking, and pavement made up 39% of total land area in the region.
Amélie Davis of Purdue applied a similar methodology in a 2010 study, which examined parking lots in Tippecanoe County, Indiana, which is home to the university. Davis and her coauthors estimated that 0.44% of the county and 6.57% of the county’s urban areas were covered in parking lots.
And, in July, Eric Scharnhorst of the Research Institute for Housing America published a study on parking in 5 American cities, finding that there are some 2.2 million spaces in Philadelphia, along with 1.8 million in New York City and 1.6 million in each Seattle and Des Moines.
The two genders of parking
Although the research may suggest that parking footprints are comparable across cities, that is not actually the case.
Cities have taken markedly different paths, as Bryan Blanc et al demonstrated in their 2014 analysis of parking and buildings in six cities. Whereas Arlington, Berkeley, and Cambridge, Hatford, Lowell, and New Haven all set aside comparable amounts (2.2-8.9%) of their downtowns to surface parking in 1950, this changed considerably by 2000. While the former three cities followed a parking-light approach, pricing parking and promoting alternative forms of transportation, the latter three adopted a more conventional approach, emphasizing ease of access for drivers. As a result, parking infrastructure made up just 5% of the parking-light cities by 2000, while it had increased to 15% in the conventional cities.
I think we can guess which of these two camps Cleveland falls into. A number of writers and commentators (including me) have lamented the ubiquity of parking in this region. The consequences of this outcome are widespread and severe. In addition to the well-established environmental costs, parking also drains the public coffers and limits economic activity.
A 2012 paper from Christopher McCahill and Norman Garrick of the University of Connecticut concluded that each unit of land area converted to parking reduces the number of people that can live and work in a city, which directly limits economic output.
Remember that parking tax I wrote about [checks notes] 32 months ago?
Given these costs, I proposed in 2016 that Cuyahoga County should levy a tax on the total surface area of parking lots and use the revenue to fund public transit. Unfortunately, because I lack both the skills and the patience of the researches I cited earlier, I was unable to estimate the total land area covered in surface parking in the County, so I could not calculate the potential tax revenue that such a tax would generate. And so this proposal sat moribund for more than two years.
That changed in July of this year, however. I was playing around with the 2010 sprawl index data for the Census tracts in Northeast Ohio from Smart Growth America, and I decided to look up a tract number on Google. I don’t remember which tract it was or what I hoped to find, but one of the first results took me to the Cuyahoga County Planning Commission’s Urban Tree Canopy Assessment from 2014.
I was familiar with the assessment and had read through it in 2014, but I had not gone through the various spreadsheets that the Planning Commission generated along with the report. It was one of these spreadsheets – the tree canopy data by Census tract – that my search had taken me to.
Identifying an overlooked datapoint
As I looked through the data, I noticed a column titled “Other Paved SQFT.” Given that there was another column devoted to rights of way, I wondered if this column could serve as a proxy for parking. I asked staff at the Planning Commission, who confirmed that this does provide data on parking area; however, they explained that, because it includes all property types in the County, it would also include things like residential driveways and patios.
Below is a map showing the total area (in square feet) of “Other Paved” land in each Census tract for Cuyahoga County. In total, there are nearly 1.8 billion square feet, which is equal to 64.4 square miles of paved surfaces. Solon is home to the Census tract with the most pavement (29.6 million square feet), while Cleveland’s tract 1118 (near East 79th and Superior) has just over 563,000 square feet of pavement.
Other paved surfaces cover 14.2% of the total land area in Cuyahoga County, with a range of just 2.1% for a tract in Gates Mills to 40.4% in North Randall.
Isolating commercial parking infrastructure in Cuyahoga County
However, as I noted, this is not a perfect proxy for parking. In order to get that, I went to the County GIS Department’s open data portal (which is awesome), where I was able to download a spreadsheet containing data on every parcel in Cuyahoga County.
Initially, I limited the parcels to just those with a land use description that include “parking.” This separated out just designated parking lots, such as those which comprise the notorious parking crater in the Warehouse District downtown.
These parking lots take up more than 14.1 million square feet of land; given that there are over 11 billion square feet of land in the County, these parking lots account for just 0.13% of total land.
I have long wondered just how much of downtown was surrendered to these massive parking lots, so I pulled the data for those parcels. These parking lots span 6.6 million square feet, or 11.5% of downtown Cleveland.
But this, again, seemed inadequate to me. While large surface parking lots are the worst eyesores, they are not the only type of commercial parking. There are thousands of other parcels that have on-site parking of some sort that are not classified as parking lots, per se.
To get a commercial parking inventory that I considered satisfactory, I took the complete list of parcels and removed all those classified as “residential,” a category that largely captures single-family homes or standalone multi-family residences. I opted to retain apartment buildings, however, because they typically include larger, dedicated parking infrastructure.
When I separated out these non-residential parcels, the “other paved” land totalled just under 940 million square feet (33.7 square miles), or 8.5% of the County. If it was incorporated as a municipality, commercial parking would be the second largest city in Cuyahoga County, supplanting Strongsville.
Below is a map displaying the square footage of commercial parking by zip code.
The literature suggests that you can estimate the number of off-street parking spaces by dividing the total area by 350 square feet. Accordingly, I estimate that there are some 2.69 million commercial parking spaces in Cuyahoga County, or 2.14 spaces per resident (based on 2017 population estimates). That also equates to just under 3 parking spaces per passenger car registered in the County, using Ohio BWV data.
Overall, commercial parking makes up 8.5% of the County’s total land area. Those shares are 14.8% and 32.8% in the City of Cleveland and downtown Cleveland, respectively. This places Cleveland squarely in line with the cities that Blanc et al classified as having followed the conventional approach to development.
Housing for cars instead of people
We can also use these data to compare the supply of parking – which, if you think about it, is essentially housing for cars – to the actual supply of housing in the County.
Let’s juxtapose the total area of commercial parking to the total square footage of buildings for apartments, residential properties, and senior housing.
Combined, there are some 677.6 million square feet of housing in Cuyahoga County, which is equal to 539 square feet per resident. In turn, there are 753 square feet of parking per resident. In other words, we have devoted 1.4 square feet to housing for cars for each square foot of housing for people in this County. This disparity is even worse in the City of Cleveland, where we provide 1.7 times more area to house cars than to house people.
The high opportunity cost of free parking
Blanc et al also identified the opportunity costs of parking infrastructure by comparing the tax revenue per square foot for buildings and parking in each of the six cities they analyzed. They found that, on average, each buildings generated 6.1 to 21.5 times more tax revenue, per square foot, as surface parking lots. This ratio was significantly smaller – just 1.4, on average – for parking decks.
Although I do not have data on tax receipts readily available, my spreadsheet does provide the assessed value of each parcel. Thus, I compared the assessed value per square foot of parking infrastructure (lots, decks, and both) to different forms of land uses (apartments, commercial retail, entertainment venues, offices, and residential properties) in downtown Cleveland.
While parking is worth just $0.26 per square foot ($0.22 and $2.37 for lots and decks, respectively), the alternative land uses are valued at $3.65 per square foot (with a range of $0.22 for commercial retail to $17.90 for entertainment venues). Therefore, the average building in downtown Cleveland is worth 14.2 times more, per square foot, as parking infrastructure. But, as with the cities in Blanc et al, this ratio is driven almost entirely by surface lots, which are 16.9 times less valuable than buildings.
So about that parking tax
I just realized that I’m nearly 2,000 words into this post, and I still haven’t addressed the central question that inspired this entire exercise – how much revenue could a surface parking tax produce?
While it would be easily to just multiple the total area of commercial parking in the County by a given tax, it’s not quite that simple. There are a number of properties that are tax exempt, whether they’re owned by the government, a nonprofit organization, or a religious entity.
To identify the total area of taxable parking, I culled all properties that the County lists as tax exempt. This shrank the population of commercial parcels by 41.7%, leaving 29,276 properties with 657.3 million “other paved” square feet. If we assessed a tax of $0.10 per square foot, it could raise $65.7 million for GCRTA, more than double the amount I initially speculated.
To put this into perspective, GCRTA reported total revenues of $286.1 million in 2017 (page 23). A surface parking tax of $0.10 per square foot would equal 23% of this total and would raise nearly 42% more than GCRTA brought in from passenger fares last year.
I freely admit that these numbers are not perfectly accurate. The parcel data I used dates back to 2010. It does not reflect that fact that several surface parking lots have been converted to more beneficial uses, such as The Quarter development on West 25th and Detroit. But it also does not include sprawling developments with abundant parking that have opened recently, like Pinecrest.
Despite these caveats, I feel that my inventory represents a decent summary of parking in Cuyahoga County, at least to a first approximation. By reserving so much land for parking, we have handicapped ourselves by facilitating ongoing sprawl and crowding better and higher uses of our land.
Addressing decades of poor land use choices and sprawl-based development isn’t going to happen overnight. But quantifying the scale of the problem is an important first step.
Someone on Twitter pointed out that it’s unwise to tax developers for building parking infrastructure when municipalities force them to build it through the imposition of mandatory parking minimums. That’s completely correct, and it was an oversight on my part not to mention that I believe we should tie the levying of this tax to the elimination of such parking minimums. Multiple cities (including our Rust Belt brethren in Buffalo) have already done so or are in the process of doing so.
This does open up a certain incongruity, though, as I would like to see Cuyahoga County levy this tax, but it is entirely up to the 59 cities, townships, and villages that make up the County to set land use and zoning requirements.
There are various ways to address this mismatch; for instance, Cuyahoga County could set aside a portion of the tax revenue (say the first X million dollars) to send back to municipalities for upgrades to transit and parking infrastructure in their jurisdictions. It could then use that funding as a carrot to push those municipalities to eliminate parking minimums.
Alternatively, forward-thinking municipalities (hopefully that would include the City of Cleveland) could act on their and pass their own ordinances taxing surface parking and rescinding parking minimums. Cleveland is home to 190.7 million square feet of taxable commercial parking, so it could raise nearly $20 million on its own. Cleveland Heights, Lakewood, and Shaker Heights could get $525,500, $516,800, and $282,500, respectively.