Which New York City Areas Have the Most Pedestrian Foot Traffic?

One of the key ingredients to a successful real estate development project is foot traffic. This is particularly important in the retail sector. Physical stores need people to walk in and buy stuff (obviously), so the more people walking by, the better.

Data regarding pedestrian foot traffic can be extremely valuable in helping to predict which locations and neighborhoods provide the best future prospects based on past trends.

With that in mind, did you know that New York City performs bi-annual pedestrian traffic counts at 114 specific locations across the five boroughs?

Did you know that they have been doing this in a standardized fashion since at least 2007?

Did you know that they post their data online for anyone and everyone to see and analyze?

I didn’t this morning. I do now. Here’s a quick peak at what I built from this data:

Change in pedestrian traffic across New York City from 2007-2015. Note that the 69% in the legend should be negative 69%.
Change in pedestrian traffic across New York City from 2007-2015. Note that the 69% in the legend should be negative 69%. Light red and dark red dots show traffic decreases.

As I’ve discussed before, New York City has done an amazing job aggregating and releasing their data. The more I search, the more I find. The NYC Department of Transit (DOT) is one of the many city entities that partakes in this open data program. They update and release at least 14 different sets of data each year.

One of those sets of data is bi-annual pedestrian counts for 114 different locations. Click here to view this data on the NYC DOT’s website.

About the data

These are the key characteristics about this dataset, according to the metadata:

114 locations, including 100 on-street locations (primarily retail corridors), 13 East River and Harlem River bridge locations, and the Hudson River Greenway

Counts taken twice a year, based on ITE recommended dates: May & September

Counts conducted on one midweek (Tue/Wed/Thu) day and an adjacent Saturday

Counts conducted from 7-9am, 4-7pm on weekday, 12-2pm on Saturday

Screenline counts conducted on sidewalk mid-block (or mid-bridge), on both sides of street where applicable

The dataset is available in Excel and as a Shapefile, so we can map whatever we find. NYC DOT also provides us with a map so we can see the different locations they track:

NYC Pedestrian Traffic Count Locations


My Analysis

First of all, please be aware that I only analyzed the September morning data for the sake of expediency. I did not analyze evenings, May, or weekends, but the analytical process I used for September morning data can just as easily be applied to all the other segments. When I get more time I will post an update.

Now, after playing with the data and lining things up as necessary, I was able to build an intuitive and useful view:

Pedestrian traffic across New York City on select September mornings.
Pedestrian traffic across New York City on select September mornings.

This alone gives you a great overall picture and enables you to do some quick analyses.

I sorted the list by which locations in New York City have the most pedestrian traffic today. By that metric, West 34th Street, Seventh Avenue, East 42nd Street, Chambers Street, and Sixth Avenue are the top five most trafficked locations in the morning. That probably doesn’t surprise you. Midtown and Downtown are have very heavy foot traffic in the mornings.

I mapped that information across all of New York City. The red dots are the top five locations noted above. The green dots show everything outside of the top five. Darker green means heavier traffic.

Here it is for you to play with:

Back to the table above. On the right, I calculated compounded annual growth rates (CAGRs) from 2007 to 2015. With that metric you can see which locations have grown in traffic.

The teaser I posted at the top of this article shows the CAGRs. Here’s the full map:

Two things jump out to me:

  1. All of the East River Bridges have increases in foot traffic. The Brooklyn Bridge, the Manhattan Bridge, the Williamsburg Bridge, and the Queensboro Bridge all have dark green dots. You see the same pattern if you look at the crossings between The Bronx and Manhattan:
    Pedestrian CAGRs from 2007-2015 for Manhattan-Bronx crossings.
    Pedestrian CAGRs from 2007-2015 for Manhattan-Bronx crossings.

    Since we’re looking at morning data, this likely shows that more people have been commuting on foot from the Bronx, Brooklyn, and Queens to Manhattan (or in the other direction). Either way, I think it supports the notion that the outer boroughs are growing and will likely continue to grow. That said, we shouldn’t overestimate based on these maps. The magnitude itself is not that much. Adding car and bike data would certainly be a valuable next step

  2. There are a number of decreases in Downtown and Midtown Manhattan. For example, Trinity Place in the Financial District has decreased at a CAGR of -69%. Wall Street and Broadway (FiDi) have decreased by 19% and 13%, respectively. There are a few decreases along 42nd Street and even West 34th Street (Penn Station) is showing a decrease of 1%.If you look at the detailed data above, I think there are a few different explanations. The West 34th Street data point appears to be an outlier. The trend across the years was steadily upwards, but it dropped significantly in 2015. If you look in the data, you’ll see that the 2015 point was measured on September 9th, but the earlier points in 2007 and 2008 were measured on September 25th. I wonder if the closeness of that measurement to September 11th had anything to do with the decrease.The city should try and review its processes to get more data samples and account for seasonal factors.Trinity Place, on the other hand, showed a pretty steady decrease across the years. There could be other factors at play such as construction and subway closures, but the trend looks clear.

Concluding Thoughts

I only analyzed a fraction of the data available and was able to build some interesting models about how people move throughout the city and how that movement has changed over the years.

With more time and thought, surely additional insights could be gleamed from this dataset alone. And if the city worked to expand its data collection practices, this source of information could become extremely valuable from real estate and urban development perspectives. Imagine if this information was collected daily and on a more micro level.

I wouldn’t want somebody counting people on every corner every single day and there would be privacy concerns, but technology should be able to do a lot of the legwork.

At the end of the day, this is a solid source of information that has the potential to get much better. I’m glad I found it.

Which Zones Have The Tallest Buildings in New York City?

Last week, in my post about the correlation between allowable FARs and building heights, I mentioned that I would be delving deeper into which zones are home to the tallest buildings. That’s what I will be exploring in this article.

To the data!

The starting point, as is so often the case, is PLUTO data. I will be using the same data that I referenced in last week’s article. The data covers the entire island of Manhattan.

If you’ve read my last article, you already know that I’ve done some calculations about building heights by allowable FAR. I expanded that model to include zones and then sorted by building heights in order to figure out the tallest zones.

Here’s what I found:

Which zones have the tallest residential buildings in New York City?
Which zones have the tallest residential buildings in New York City?

This table shows zones on the Y-axis and allowable FAR along the X-axis. As expected from my last post, when we sort the zones in order from tallest to shortest (as I did above), we get a trend that moves up and to the right.

This trend exists because allowable FAR has a positive correlation with building heights. The higher the FAR, the taller the building, on average.

The column on the far right is added for reference as a measure of magnitude. It is the total number of buildings within a given zone in our sample.

This is how you read the chart. The first zone, C5-2.5, has a maximum allowable FAR of 10 (from the top row), an average building height of 32.0 stories (the green cell), and contains 9 residential buildings (the column on the right).

Next, I’m going to throw this data into Carto and do a visual spot check. Which 9 residential buildings are in C5-2.5? Does this make sense?

The residential buildings in zone C5-2.5.
The residential buildings in zone C5-2.5. (Color scale dictates height.)

If you know New York City real estate, the big red zone in the middle of the map will stick out like a sore thumb. That’s 432 Park, the 90-story residential skyscraper home to some of New York’s most expensive apartments. The dark orange building highlighted to its left is Museum Tower, a 53-story residential tower at 15 West 53rd Street. And to the south, just west of 3rd Avenue, is The Metropolis, a 48-story luxury rental building at 150 East 44th Street.

So it looks like we’re on to something. Simply modeling average building height by zone has enabled us to quickly isolate some valuable properties and create a hypothesis that zone C5-2.5 is great for tall buildings.

Warning: Digression Ahead

Now I want you to take a closer look at that map. Squint at the bottom left of the map between Park Avenue & Lexington Avenue and you’ll notice a very faint yellow rectangle. That’s 114 East 40th Street. And it’s only 9 stories tall. Weird.

In zone C5-2.5, we have 432 Park standing 90 stories tall, dominating the New York City skyline, and we have 114 East 40th Street rising just 9 stories.

Why? What other factors could create such an enormous disparity in building height?

Both buildings are C5-2.5 and both buildings are in the Midtown Special Purpose District (MiD). Zone clearly is not the only important factor in determining height.

432 Park’s lot area is about 10x the size of 114 East 40th Street’s lot area, but 432 Park has a building area of 745k SF whereas 114 East 40th has a building area of 26k SF — a 30x differential!

There is the added benefit of 432 Park being along a wide street, as explained in my post about setbacks. 114 East 40th Street was also built in the 1920s when the city was generally shorter.

But I think the key here, as reported by the New York Times in 2013, is air rights. 432 Park came with 115k SF of additional air rights before it was built. That enabled it to scale to great heights.

Unfortunately, there is no information about transferred or additional air rights in PLUTO data and I have not found a good source for this type of data, so we will have to live with the fact that we cannot account for air rights in our analyses automatically.

That said, we can still draw valuable conclusions. Zone C5-2.5 plus transferred air rights can lead to huge buildings. It’s something worth looking into for the other zones as well. For example, which sites are most favorable for applying transferred air rights?

Additionally, there is a lot more to building height than the zone in which a building sets and, therefore, there is a lot more to building height than allowable FAR (which is dictated by zone). Perhaps I’m inexperienced, but that’s news to me. The impact of air rights might be more meaningful than I previously thought.

Let’s get back on track.

That was a serious tangent. Valuable, but serious. Here’s the same breakdown for commercial land use:

Which zones have the tallest commercial buildings in New York City?
Which zones have the tallest commercial buildings in New York City?

It’s interesting that C6-6.5 has, on average, the tallest buildings, but it does not have the highest allowable FAR. I’m not exactly sure why that is, but I already bored all of you with one digression, so I won’t do it again.

Additionally, C5-3, C5-5, C6-6, C6-7, and C6-9 are all in the realm of 20-25 stories. If I were a real estate developer looking to build commercial skyscrapers, I would certainly focus on properties in these zones as a starting point.

In conclusion

This simple analysis has provided us with a foundation as to where the different zones stand in relation to commercial & residential building heights. I think it’s a quick and useful tool to help judge potential building height. On average, the zones towards the top of the list are going to be taller and inherently more valuable than the zones towards the bottom of the list. But that isn’t a definitive rule and it isn’t a suggestion that proper due diligence need not be performed.

As we saw with our comparison of 432 Park and 115 East 40th Street, zone is just one piece of the pie in attempting to figure out the key drivers behind building heights. I will continue to explore these drivers and use them to identify value where possible.

Correlation Between FARs and Building Heights in New York City

In my last post, where I modeled the real estate growth potential of Manhattan, I alluded to Floor-Area Ratio (FAR) as a metric not just for square foot growth potential, but also for height. I showed how FAR is related to height, but explained that there are many other constraints on height in the zoning code, so it isn’t a perfect metric.

That led me to wonder if there is truly a correlation between building height and the FARs allowed by the zoning code. I assumed there was (because the zoning code states as much), but there’s never any harm in checking the data. It’s always possible that the actual outcome of a policy deviates from the intended outcome.

So I dove into New York City’s data (courtesy of PLUTO) and, to no surprise, found that a strong correlation does exist.

As Allowable FAR increases, average # of floors increases too.
As Allowable FAR increases, Average # of Floors increases too.

The above chart plots, for both residential and commercial buildings, the average number of floors per building (numfloors in PLUTO) within the different possible allowable FAR buckets as dictated in the zoning code (residfar for residential & commfar for commercial).

You can see that the graph trends upwards. Higher allowable FARs have, on average, taller buildings. This makes perfect sense because you can’t build a tall building if you don’t have a lot of floor area available to utilize.

Of course, this general outcome was expected. The New York City Zoning Resolution purposefully puts high FARs on areas where it wants to allow taller buildings.

But it’s interesting to note that the correlation is not perfect. Look what happens if we segment the commercial building data into three buckets:

The data for commercial buildings shows three distinct buckets.
The data for commercial buildings shows three distinct buckets.

You can see that zones with maximum allowable FARs from 1 to 3 don’t seem to grow in an orderly fashion. Zones with maximum allowable FARs from 4 to 8 all have about the same average height. I would think a zone with an FAR of 8 would have much taller buildings than a zone with an FAR of 4, but that doesn’t appear to be the case. From this analysis, I can’t answer why, but it’s worth looking into.

What’s really interesting to me is that building height growth takes off going from FARs of 8 to FARs of 10 and up. On average, an FAR of 10 spawns buildings twice as tall as an FAR of 8, even though that’s only a 25% increase in FAR. There have to be some other factors encouraging this type of growth.

This is all important to know if I’m a real estate developer. At face value, an FAR of 8 doesn’t sound too different from an FAR of 10, but the data shows that there can be a huge difference in terms of building height. Who knew? I guess it’s because of tower regulations, but I haven’t looked into those too much (yet), so I can’t say definitively.

Gotta run!

Summer intensives just started at NYU, so I’m short on time this weekend and have to cut out here, but know that there is more to come on this subject. In my next article, I’ll be diving into further detail about which zones produce the tallest buildings and what that means for real estate developers.

For reference, the data I used spans the entire island of Manhattan:

The data for this analysis contains all of Manhattan.
The data for this analysis contains all of Manhattan.