Post #459: Park once, shop many and mixed-use trip reduction.

Posted on November 15, 2019

Google Earth view of the Giant Food shopping center parking lot.  Maple Avenue is at the top of the image.

When I was first introduced to MAC zoning, I assumed that, somewhere, somebody had done some hard analysis of how MAC was supposed to work.  Much later, I came to realize that often was not true.  Much of the analysis never went much beyond using current urban-planning phrases, combined with the hope that somehow those things would happen here.  There really wasn’t any analysis of how, exactly, that would work on Maple Avenue.

I could list a few, but if you read this blog, you can fill those in on your own by now.  If you want to see one, look at Post #302, on “destination shopping”.  The last graphic in the post is a stark contrast between actually creating a true destination shopping district (in this case, the Mosaic District), and just saying those words about Maple Avenue.

In this post, I’m going to drill down into two related concepts:  “Park once, shop many” and “mixed use trip reduction”.  These are routinely touted as advantages of mixed-use development.  My point is that if you actually look at the details, on Maple Avenue, you quickly realize that there’s not much there, there.  Quantitatively, the impact of these factors, on Maple, is apt to be quite small.

What are these concepts, and why do they belong in the same discussion?

The are both about substituting foot travel for car travel.  That can occur when shops are near one another (park-once-shop-many) and when homes are near shops (mixed use trip reduction).

At the 11/9/2019 Town Council work session (Post #450), I once again heard about “park once, shop many” and “shared parking” as desirable features of MAC.  This is the idea that you park your car and walk to several different stores, rather than driving to each store separately.

The traffic consultants also gave a vague reference to a related concept:  mixed-use trip reduction.  Broadly speaking, people who live directly on Maple will be more likely to walk to Maple Avenue retail, rather than drive. Accordingly, when traffic engineers estimate the additional car trips from a new mixed-use building, they sum the data for each piece of it (as if each piece of it were stand-alone), then reduce that some.

The reason I put these two concepts together is simple.  At root, they are both about substituting walking trips for car trips, due to the proximity.  In the first case, you had to drive to Maple and park.  In the second case, you already live on Maple (and you might bike, in addition to walk).

OK, these are real things, no doubt.  But the question is, do they matter?  Quantitatively, how big are these effects likely to be, on Maple Avenue?

Less ADDITIONAL traffic.

On this mixed-use trip reduction in particular, I feel the need to be very clear about one word that Town staff invariably forget to use:  ADDITIONAL.  I don’t know if they drop that word with the intent to mislead or not.  But the fact is, their discussion is always misleading.  Invariably, they will say that mixed-use development results in less traffic.  That’s wrong.  Mixed-use development results in less ADDITIONAL traffic than you would get from an equivalently-large set of single-use structures.   

The key point here is that mixed use on Maple will not result in less traffic.  It will not literally reduce the number of cars on Maple, compared to what’s there now.  It will result in less of an increase in traffic than you would predict from building new buildings that are the same size as these big MAC buildings, but building the housing and retail components to be separate, independent buildings.

But “less traffic” is exactly what Town staff were saying, more than a year ago.  And that phrase got tossed out again, in the last Town Council work session.

When I first got into this, I found that discussion to be so head-spinningly illogical that that I had to write it all down to get straight what they actually meant.  That resulted in one of my first posts here, on “walkability”.  And it all boiled down to Town staff omitting the word “additional” when they talked about mixed-use buildings and traffic.

Let me try to be completely clear.  Suppose you have two vacant lots, one on Maple, one half a mile away.  If you build a new shopping center on Maple, and an apartment complex half-a-mile away, you’ll generate a certain amount of total new traffic (to both of the single-use buildings).  By contrast, if you stack the apartment complex on top of the shopping center, you’ll generate slightly less total new traffic, because the apartment dwellers don’t have to drive to the shopping center.

Empirical estimate:  Mixed-use trip reduction, 10%

If nothing else, this analysis should convince you that estimating traffic generated by new development is not a science.  At best, it’s more-or-less a set of rules of thumb, with some self-reported, sample-of-convenience data to back them up.

When I went to look for estimates of a typical mixed-use trip reduction factor, what I quickly learned is that estimates vary wildly.  In part, that’s because a mixed-use development could be anything from a handful of apartments over a store, to more-or-less a self-contained small city.  But that’s also because methods and particularly data used for the estimates also vary wildly.

I did pick up a few rules of thumb, one of which the most important one is that size matters.  Bigger developments and denser population provide more opportunity for individuals to satisfy their shopping/dining needs within walking distance.  Smaller and more spread-out mixed-use developments provide correspondingly fewer opportunities, and so should have a much smaller mixed-use “discount” applied to the standard trip generation estimate for stand-alone buildings.

What we’re going to get on Maple will be small mixed-use developments, as these things are routinely discussed.  For one thing, there’s only one road, not a bunch of city blocks in proximity to one another.  For another, there’s a four (or is it five?) floor limit.  And the total area in question is just not that big to begin with.

If you want to get a flavor for the rule-of-thumbness of this, realize that state-of-the-art software for doing the mixed-used trip reduction calculation is a spreadsheet.  You can download that yourself from the Institute of Transportation Engineers.   No slam against spreadsheets, but clearly this is just a handful of lookups, and a little calculation.  Fill in your best guess for what your mixed-use development looks like, and it’ll calculate the trip reduction.

My only problem with the ITI spreadsheet is that I quickly realized that I didn’t know enough of the jargon to be able to fill out the spreadsheet.  So I went looking for somebody who could dumb it down for me.

When I want a simple, straightforward answer about a traffic engineering question, I go to this guy, Mike on Traffic.  He has been my long-standing source in this area mostly because he writes clearly and eschews jargon.

And, as it turns out, a lot of traffic engineers don’t even go to that spreadsheet level of sophistication.  On that particular page, he gives gives four typical estimates for mixed-use trip reduction including his own personal rule of thumb:

"Use a 10% across the board reduction: I typically use this blanket reduction as do most of the traffic engineers in Minnesota. Given the above datasets, I think it’s reasonable if not slightly conservative. It’s also quick to use."

 And so, there you have it.  Figure out how much total additional (new) traffic you’d get from building a brand-new shopping center and, separately, a three-story apartment building.  If you then stack the apartment building on top of the shopping center, you can take 10% off that estimate of additional (new) traffic.

That is the much-vaunted mixed-use trip reduction.  But if you think about it even a little deeper, you might eventually figure out that my example was unrealistic.

If you have two redevelopable lots on Maple, what you’re going to get with MAC is not EITHER a new-shopping-center-plus-separate-new-apartment-complex OR the mixed-use-building-with-one-stacked-over-the-other.  If you have two redevelopable lots, what you’re going to get is two new mixed-used buildings.  You’re going to get two new shopping centers and two new apartment buildings.

If you figured that out, you get an A+, because that’s the empirically important bottom line about MAC redevelopment.   The overwhelming impact of MAC mixed use is to add a whole lot more stuff to Maple Avenue, period.  That new stuff is going to generate new traffic.  The much-touted mixed-use trip reduction is more-or-less rounding error by comparison.  All it tells you is that your best guess at ADDITIONAL traffic will be slightly less than you would naively estimate from simply summing the traffic for each portion of the building.  How much less?  Something like 10%.

Empirical estimate:  Park-once-shop-many/shared parking?  Nobody knows.

This section is going to be fairly short because I could not find even one estimate of the extent to which shared parking, or park-once-shop-many, reduces car trips. 

Everyone seems to agree that, if done right, shared parking can reduce the total amount of required parking spaces.  And that, not traffic, appears to be the main advantage.  A classic example of that occurred right here in Vienna with the proposed Bear Branch Tavern.  This is going to be in the first floor of an office building, and they made the argument that they could reasonably share parking between the restaurant and office uses, for the simple reason that restaurant does most of its business on nights and weekends, when the offices are closed.

But in the context of MAC and redevelopment, all that shared parking means is that you can reduce total parking, and so you can build even more new housing and retail on Maple Avenue, in the same amount of space.  From my perspective, that’s not coming ahead.

But in terms of the number of car trips avoided by shared parking, as far as I can tell, nobody has any sort of empirical estimate.  There might be one out there, but if there is, I could not manage to unearth it.  But that leads me to suspect that the net effect of shared parking on the number of car trips, and particularly on the number of vehicle miles, must be fairly small.

Now, when I first described “shared parking”, if you said “oh, you mean, like a shopping center or a mall”, then you’ve got the concept.  Although I’m sure the people who push that as some unique aspect of MAC would blanch at that.  But let’s run with that.  In practice, as far as I can see, shared parking is simply an attempt to retrofit the shopping center concept on some otherwise independent set of merchants.

Since there doesn’t appear to be any hard data, let me use the Giant Food shopping center for a little thought experiment on shared parking.  First ask yourself this question:  When you shop at the Giant Food shopping center, how often do you go to more than one store?  (Versus, say, just picking up groceries, or just filling a prescription.)

Every time you do that — and just walk from one store to the other — you have avoided what would otherwise be a car trip.  You have taken advantage of the shared parking, by doing a multi-store trip.  You have parked-once-and-shopped-many.

Call that a multi-store trip.

Now imagine that shopping center parking lot was divided up, so that each store had its own slice of the lot, you had to go back out to Maple to get from one slice to the other, and you could not leave your car in front of one store and shop at another.  (When you think about it, this is more-or-less how much of the north side of Maple is.  While the south side is mostly shopping centers, the north side has some long stretches of stand-alone merchants, each with their own parking.)

With this imaginary divided parking lot, that multi-store trip now requires an additional car trip on Maple.  But note that it’s a very short trip.  You would only have to be on on Maple for a few seconds.  You only travel a fraction of a “vehicle mile” on Maple.  You only occupy a fraction of a “vehicle minute” in Maple.  You don’t really use Maple all that much.  Certainly not compared to a commuter passing through, driving down the entire length of Maple.

And so, at this point, I’m just guessing, but my guess as to why I couldn’t find an estimate of the impact of shared parking on traffic congestion is that it’s not very large.  Sure, shared parking — as in a shopping center — lets you avoid some car trips.  The trips you avoid with that are very short distance, very brief duration car trips.  And, arguably, in the typical case, those very short, very brief trips add little to the congestion on the adjacent streets.

The final thing to bear in mind is that shopping centers — where one economic entity owns the entire row of shops — carefully plan the location and mix of tenants.  (At least, healthy ones do.)  Those choices are designed to maximize the synergy across tenants, and so maximize the number of multi-store visits that the average shopper makes.

Classically, you had “anchor” stores at either end, preferably a grocery store for one, and (historically) a department store for the other.  Other infill tenants between the anchors were chosen to provide a complementary array of other frequently-needed retail services (bank, dry cleaner, drug store).  And the remaining infill was specialty retail.  It was all designed to maximize synergy, minimize retail conflict.

Obviously, that anchor-store model is not the only one.  The former Magruder’s shopping center used to fit that model reasonably well.  But now it’s more-or-less a food court.  They gave up on the anchor store concept entirely, and instead provide a wide array of fast-casual dining (plus a few other services).

My point here is that true shared parking (where each shop and lot has a different owner) is going to be much more of a hodge-podge than a shopping center (where one owner coordinates the tenants to maximize synergy.) Instead of stores chosen for synergy, it’s going to be stores that happened to be close to one another.  And that’s going to mean less opportunity for park-one-shop-many than you find in a typical, well-planned shopping center.

So, as the final thought experiment, take a look at the stand-alone businesses along the north side of Maple.  Imagine that you connect the parking lots on any given block, and can park at any of of them and shop the rest.  Do you see a lot of opportunity for park-once-shop-many among the existing hodge-podge of stand-alone shops?  If not, then park-once-shop-many, even if it could be implemented, would likely have a negligible effect on traffic.

Let me just take my end of Maple.  If I pretend that all of the shops were open, I’d have a convenience store (Wawa), restaurant (former Joe’s Pasta), coffee shop, bridal shop, orthodontist, bank, and an eye doctor.  In fact, the only “synergy” retail we have is the bank, and then some small shops that are, in fact, already part of a shopping center (e.g., a barber shop).  Nothing at all like the classic grocery-drug store-dry cleaners of a traditional shopping center.

For myself, at my end of town, I’m not seeing a lot of multi-trip synergy among the individual merchants who happen to have located here.  I suspect that’s how it is across most of the physically-adjacent independent merchants on Maple.  And so that, combined with the fact that the trips avoided by park-once-shop-many, leads me to think that this concept, if it could be applied to Maple, would have a negligible impact on traffic.

An extended aside on transit

There is yet a third advantage attributed to mixed-use and traffic, but I’m going to ignore it for this discussion.  That other advantage is that you might get people to substitute bus trips for car trips.   (We can say “public transit” instead of bus, but the only public transit on Maple is the bus.) 

The point is that when you place high-density housing a bus route, you put more people within easy walking distance of a bus stop.  (More than you could with single-family homes).  For example, 444 Maple West will hold about 250 adults who could walk 400 feet to catch a bus to Metro.  Arguably, that’s more people than live within 400 feet of a bus stop in the entire Town of Vienna now.

I’m going to ignore that, in this post, for two reasons.  First, in my experience, that argument is about commuting to work, not shopping.  In the entire time I’ve been riding the Connector bus in Vienna, I have only once seen somebody get on holding shopping bags.  Second, the first-order effect there is more about self-selection than anything else.  That is, people who have made up their mind to use public transit will choose to live near transit.  To a first approximation, mixed use doesn’t increase total transit use, it just redistributes it.  (If it did increase bus use, we’d be unlikely to see steadily declining bus ridership in this area, with all the mixed-use developing going in.)  But high-density development has a beneficial secondary effect, in that it concentrates transit users in one place, which makes buses much more economically viable.