Friday, August 27, 2010

Parking Lots -- Too Big and Too Little

The planned Tavistock development is slated for the upper area of this photograph, to the left of Apopka-Vineland Road and north of the Cascades Publix shopping plaza.  The photo captures a typical day when the half of the parking lot closest to Apopka-Vineland Road (and furthest from the Publix) sits as empty asphalt.  In contrast, St. Luke's Methodist Church, across the street, relies on grassy areas for overflow parking.   

Tavistock is requesting a waiver to remove several hundred parking spaces required under the Orange County Code at a planned development at the southwest corner of Apopka-Vineland and Conroy-Windermere Roads.  Tavistock is justifying the request on the basis that calculations for required parking spaces are based on single use, whereas Tavistock is planning multiple uses that will experience peak hours at different times. 

For example, Tavistock's planned fitness center will generate the most parking in the early morning and evening whereas the bank and office building will generate the most parking during business hours.    Even during the noon hour, when fitness center use increases, total parking generation will not exceed 1,000 spaces, according to a shared parking analysis by the planning firm Glatting Jackson. 

These business should share parking spaces.  To save unnecessary construction costs, every proposed mixed use development should conduct, and the County Code should recognize, a shared parking analysis. 

The County's voluminous parking requirements are based on annual peak demand--for example, the Sunday before Christmas for retail development--which leaves huge expanses of half-empty asphalt 90% of the time.  Our off-street parking requirements are generally too large and denigrate the built environment. 
Typical half-empty parking lot in front of strip shopping center on Hiawassee Road.

Winter Park K-Mart Plaza.  Typical parking patterns with huge expanses of empty asphalt. 

Developers pay for all the extra, unused asphalt, which drives up rent, which increases the cost of goods and services to consumers.  We pay this hidden parking fee everyday.  If the County grants Tavistock the waiver it seeks, it will avoid the cost of building a structured parking garage, which could cost about $10,000 per parking space.  Such a garage would drive up rents and consumer costs. 

I am the first to concede notable instances when we can't find enough parking.  I often find a packed parking lot at The Fountains on Sand Lake Road on Friday and Saturday evenings, when its myriad restaurants (all the same land use) are doing the most business at the same time.  Winter Park Village, on 17-92, historically had insufficient parking at peak shopping and eating times. 

In most instances, however, when we say we can't find parking, we can't find convenient parking within a few hundred feet of our destination. 

ITE Parking Generation Calculations--Voodoo Mathematics

Most local governments base parking requirements on mathematical formulas published by the Institute for Transportation Engineers ("ITE").  For example, a fast food restaurant with a drive-through window must have 9.95 parking spaces per thousand square feet.  Consequently, most suburban 3,000 square foot McDonalds and other fast food restaurants nationwide will have 30 parking spaces--regardless of the availability of transit, walkability from adjacent neighborhoods, or actual experience.  

Donald Shoup, a University of California at Los Angeles professor of land planning, analyzed the ITE requirements and found them statistically indefensible.   ITE bases many of its parking demand graphs on a miniscule number of studies of limited observations in auto dependent environments lacking transit. 

Shoup sharply criticizes the parking requirements for drive-through fast food restaurants.  ITE reports average parking generation of 9.95 spaces per thousand feet.  Shoup argues this precision gives a false sense of mathematical certainty.  A closer look at the ITE graph below (from the 1987 edition of ITE's Parking Generation) shows that the 9.95 average per 1,000 square feet bears no relationship to the wide range of parking demand.  Square footage is a poor basis on which to calculate parking demand.  

ITE plotted peak parking demand on the vertical axis and restaurant square footage on the horizontal axis.  The plotted squares measure the various observations.  The diagonal line marks the average and serves as the basis for the precise 9.95/1,000 square feet ITE parking generation calculation.  Local governments adopt ITE's calculations for maximum, peak demand as minimum Code requirements.  However, the diagonal line bears virtually no relationship to the scattered observations. 

The chart above plots the basis for a regression equation. (Don't let the mathematical term scare you--read on.)  A regression equation measures predictability on a scale of 0 to 1--from zero predictability to complete predictability.  If parking demand truly and absolutely related to square footage, the plotted squares would all fall on the average line and the regression calculation (R squared) would be 1.  If restaurant square footage was a significant, but not exclusive factor in parking demand, the plotted squares would fall near the diagonal line.  However, in the chart above, the plotted squares fall haphazardly all over the chart. 

One of the largest restaurants, at over 5,500 square feet (on the right side of the graph) generated only about 20 full parking spaces at peak hour.  A restaurant more than half as small, 2,500 square feet, generated about 35 full parking spaces, tying for fourth highest on the chart.  The regression calculation on this chart (R squared) stands at an abysmal 0.038--very close to zero.  The line drawn by ITE is statistically indefensible.  In later editions of its Parking Generation book, ITE removed the R squared calculation. 

Our everyday experience confirms that square footage does not necessarily mean more parking usage. Despite voluminous passer-by traffic, most parking spaces at the Burger King at Kirkman and Colonial are empty at dinner time. A parking lot at a much smaller restaurant, the McDonalds on 17-92 in Maitland, is often packed at the same time.   

Several years ago, when Professor Shoup asked ITE to publish his article articulating these deficiencies in its Journal, ITE refused.  I would have preferred to see ITE acknowledge deficiencies and start a dialogue within the transportation engineering community about how to improve the methodology.  I would like to see ITE publish recommendations for shared parking, parking for areas served by transit, transit oriented development, and walkable communities. 

Corporate Demand for Parking Spaces

Developers contend the corporations they want to lure are driving large numbers of parking spaces.  CVS requests 75-80 parking spaces for each store location.  At the Tavistock community meeting last week,  attended by several dozen people, I asked how many people had ever seen 80 people in a CVS.  Not one hand raised. 

Walgreen's at C.R. 535 and Tilden Road in the City of Winter Garden.  Typical pattern--fifteen parking stalls used and more than forty empty.  The store would have more than sixty empty stalls if this were a CVS with the usual 75-80 parking spaces.   

However, even CVS will depart from its prototype.  The CVS in Baldwin Park has a limited number of on-street parking spaces in front and a shared parking lot in the rear, in the middle of block, hidden from street view.

Baldwin Park CVS.  Limited, shared on-street parking in front and a shared parking lot in the rear.

Many corporations have not grasped that half-empty parking lots make their investments appear economically unhealthy.  In the Pine Hills, this contributes to the perception of downward spiraling businesses. 

Strip shopping center at Pine Hills Road and Silver Star. 

In contrast to CVS, Home Depot conducted a study of its stores and, based on the findings, lowered its parking requirement from over 900 spaces to around 540 spaces per store.  Lower development costs can enable Home Depot to gain a price advantage over its closest rival, Lowes. 

Orange County staff has grasped the over-abundance of parking and no longer permits developers to pave more than 10% over the minimum requirement.  The current code hinders shared parking by requiring Affidavits.  As the planning staff works on a Unified Land Development Code, it should incorporate provisions encouraging shared parking, including payment into a fund for shared parking lots, and loosen the virtual prohibition on on-street parking. 

UPDATE 8/29: At last week's community meeting, I incorrectly identified Professor Shoup as from the University of California at Berkley.  He is at UCLA. 

Sunday, August 1, 2010

Overstating SunRail's Cost to Orange County

Mayoral candidate Matthew Falconer sent a mailer accusing his opponents of voting "to spend $1.5 billion on a rail system we the people rejected by our vote."  (An earlier mailer claimed $1.6 billion.)  His opponents did, in fact, support SunRail, but they did not vote for anything close to a $1.5 billion Orange County budget obligation.  The SunRail Interlocal Agreement, excerpted below, shows Orange County's actual share of SunRail expenditures will amount to just over $40 million over seven years.

SunRail Interlocal Agreement, p. 14.

$40 million is a lot of money--and Orange County needs to monitor to make sure it's spent wisely--but the cost is in line with expenditures for typical road widening projects.  For example, the Planning and Zoning Board voted two months ago on a $75 million road widening project for Southeast Orange County--covering only four miles.  (SunRail will ultimately span 61 miles).  The Orlando Sentinel reported on August 5 that 3 miles of road widening in East Orange County is costing $27.8 million).  For another comparison, the Wekiva Parkway to I-4 interchange will cost about $450 million (slated to come from our tolls unless Congressman John Mica secures federal funds).  No politician is making any of these  expenditures a central campaign platform.   

I'm not certain how Matthew is calculating $1.5 billion for the SunRail system.  Here's my understanding of the numbers: capital construction, right-of-way, trains, and soft costs are expected to amount to just under $600 million, according to a 2009 economic impact report:

Aecom Economic Impact Final Report, p. 4.
SunRail is budgeting $615 million for these capital costs.  In addition to that figure, the Florida Department of Transportation is paying CSX Transportation, Inc. $432  million for the 61 miles of track, right-of-way, and facilities upgrades, (with a leaseback from CSX to operate freight trains at night).  This figure includes $150 million, which CSX committed to spending for upgrading track and facilities elsewhere in Florida to which CSX will divert freight trains. 

Second Amendment to the Contract for Sale and Purchase, March 29, 2010

In addition, FDOT is building five bridges to separate the CSX line from grade level in Alachua, Sumter, and Marion Counties for $214 million--work scheduled independent of SunRail.  I total all that ($615 million plus $432 million plus $214 million) to reach $1.25 billion.  (Matthew's book on p. 169 also computes $1.25 billion). The United States Department of Transportation is covering half with funds they'd otherwise send to other states.  I can assume only that Matthew's $1.5 billion SunRail figure includes years of maintenance and operating expenses (a figure rarely, if ever included when we discuss the cost of roads and highways) and potential cost overruns (although construction costs to local governments have declined with the recession). 

[UPDATE 9 10 10: Matthew informs me the $1.5 billion figure is the total cost to all local governments over twenty years according to the FDOT.  I'll post his entire email below.]

Congressman Mica, Senator Dan Webster (before he left office), and every conservative member of the Greater Orlando legislative delegation presumably compared the cost of the SunRail to the cost of right-of-way acquisition and construction of 61 comparable miles of I-4 lane capacity--SunRail cites $2.3 billion for 30 miles of I-4--and concluded SunRail's $1.2 billion shared price tag was worthy. There's a lot of collective wisdom in that group.

The Voters' Will

Matthew claims the voters rejected SunRail. They didn't. Seven years ago when Mobility 20/20 appeared on the ballot, voters rejected a different system, on a different route (not on the CSX line), intermixed with I-4 toll lanes (derided as "Lexus lanes"), and an acceleration of the State and County's voluminous roadbuilding schedule. Taking Matthew's rationale to its logical conclusion, to uphold the voters' intent, one would also need to oppose all the taxpayer subsidized roadbuilding, too, rejected by the voters. (Not a position I would take).

Property Taxes Not Slated for SunRail

Matthew filed a lawsuit calling SunRail "unconstitutional" on the grounds that operations and maintenance after seven years would require expenditures of property taxes without a popular vote.  Matthew faces an uphill legal battle.  He will need to demonstrate that the Interlocal Agreement removes local government funding flexibility.  The Interlocal Agreement does no such thing and, further, each local government's share of debt service must come from "non-ad volorem sources," that is, not from property taxes. 

Interlocal Agreement, pp. 21-22. 

I could not find anything in the agreement that would obligate local governments to fund operations and maintenance from property tax revenues.  In contrast, shifting more financial burden to road construction and maintenance, the only option Matthew would leave us, would increase long-term pressure to raise property tax millage rates.  Orange County's 4,500 miles of roads (which could stretch to Los Angeles) and State roads don't pay for themselves.  Their ongoing maintenance, resurfacing, and ultimate rebuilding costs reach truly staggering proportions.

"...shifting more financial burden to road construction and maintenance, the only option Matthew would leave us, would increase long-term pressure to raise property tax millage rates.  Orange County's 4,500 miles of roads...don't pay for themselves"

In any event, a proposed $2.00 daily rental car surcharge--less than the cost of a "butterbeer" at Universal's Wizarding World--would provide a plausible source of funding for Orange County's share of SunRail operations and maintenance beginning in the year 2020. 

Rail Successful in Phoenix Despite Sprawling Development Patterns

I agree with Matthew that sprawling development patterns are not conducive to rail transit.  However, a local traffic engineer who returned from sprawling Phoenix a couple weeks ago told me their new light rail system is successful despite surburban development patterns.  The following video confirms ridership exceeding expectations:

Phoenix’s METRO Light Rail Takes Flight from Streetfilms on Vimeo.

Ridership on rail transit systems throughout the nation is reaching near record levels (though down from when gasoline was $4.00 a gallon), as many links in the right-hand column demonstrate.  When I worked in Philadelphia after law school in the early 1990's, friends would drop off their cars in Cherry Hill or other locations in New Jersey and take the train to their jobs in Center City.  SunRail stops furthest from downtown Orlando have considerable parking planned.  It's better to under-promise and over-deliver, but all this does bode well for SunRail's ridership, if SunRail is done right. 

Time Away From Families

For many who spend around 50 hours annually sitting in I-4 traffic, SunRail can provide a congestion-free option.

Questions for Conservatives

For conservatives pondering all this, I'd pose the following questions:

1.  Where are the family values of leaving us with no choice but to spend the equivalent of more than a work week in traffic away from our families? 

2.  Where's the commitment to providing economic opportunities for small businesses?  Opposing rail also means opposing the mixed use, small-business dominated, transit oriented development SunRail would foster.  (Light rail lines in Charlotte, Portland, Seattle, and other cities have generated billions in such development around stations.  Check out pages 198-202 of the report at this link.) 

3.  Where's the commitment to our national security by leaving us overly dependent on Hugo Chavez, the Saudis, and other foreign sources for our transportation energy needs?  Does it serve our interests to send so many billions of our nation's wealth overseas?  Does our susceptibility to oil price shocks--such as those in 1973, 1979, and 2008--advance our economic interests? 

4.  Don't we want America to be #1?   Shouldn't we have a mass transportation system at least as good as the Czech Republic's

5.  Why did Ronald Reagan make no effort to curtail massive federally-funded expansion of the Washington, D.C. Metro system into suburban Virginia and Maryland during his presidency?   

Small businesses dominate Transit Oriented Development in Orange, NJ
Roads Alone Can't Solve Congestion

Before the recession, Orange County was spending roughly $355 million each year on capital roadbuilding projects.  Our Commissioners trimmed that to around $150 million as property tax receipts and impact fee revenues declined.  The reality is that, for nearly two decades, we haven't, and couldn't afford to build our way out of congestion with roads and highways alone:

We haven't come close to keeping pace with our increasing vehicle miles traveled, even while paving over Orange County with extremely wide, high speed roads making us #1 in the nation for pedestrian danger.  Matthew claims that SunRail will take away funding for road safety.  The reality is that we devote much of our road "improvement" funding to creating awful environments like University Boulevard, shown in the photo below, which place pedestrians (you can find at least two in the wrong place) in danger:

University Boulevard, Orange County, Florida

The bottom line is that, while SunRail won't eliminate congestion, I-4 will have more congestion without it.  Matthew is correct that passing trains will delay motorists heading to I-4.  However, SunRail is cutting those delays in half by using double-decker trains

Massive Government Subsidies and Market Intervention

Matthew tells audiences the fact they drove to the event where he is speaking, instead of using mass transit, is a "free market choice."  Cultural conservative William Lind would disagree.  He wrote a thoughtful article published this month on the conservative case for rail transit.  He argues that our rail-free, auto-dependent lifestyles are not a free market choice, but rather reflect massive, decades-long government intervention in the market consisting of road and highway building subsidies.  Until the 1950's, the nation's private rail carriers flourished--and paid taxes.  However, when the Government taxes one economic activity while subsidizing a competing one, the competing one will undoubtedly prevail.  You can find Lind's interesting article at this link

Here is a video in which Mr. Lind makes these arguments:

William Lind: A Conservative Voice For Public Transportation from Streetfilms on Vimeo.

UPDATE 9 10 10: I told Matthew that, in all fairness to him, I would post his Response:

From: Matthew Falconer
To: Richard Geller

Cc: Paula Dockery
Cc: Beth Dillaha

Sent: 9/8/2010 1:59:03 PM

Hi Rick. I was sent your blog on my overstating Sun Rail cost. The $1.5 billion is the cost to LOCAL governments over the first 20 years as provided for in the FDOT analysis. Given Miami's Tri-Rail loses $87 million a year and has 5 times the rider ship I think Sun Rail will repeat that expense.

Your love for new urbanism and mass transit is allowing you to look at facts that support your desired conclusion. I only look at facts. Our nation is $14 trillion in debt. Our state has a $7 billion budget deficit. Twenty five percent of mortgages in Orange County are in default and 75% of small business are losing money. Twenty Five percent of rental car cost is already tax.

Seminole County is seeking a tax increase to pay for mass transit and Orange County will after the elections. This will take $300 million out of our small business economy. I may have lost the election but I was not incorrect about one thing; we are killing the golden goose of small business.

It is hard for insiders, lawyers, engineers and others to understand how bad the economic conditions are. It can and will get worse if we continue to add to the tax burden of consumers and small business.

Matthew Falconer
Falcon Real Estate Solutions