New York Got Louder in 2020. It Hasn't Quieted Down.

There's a running joke about New York being the loudest city in America. New Yorkers complain about the noise constantly - and, helpfully, they call 311 to do it. That gave me an excuse to pull every noise-related 311 complaint the city has recorded since 2010 and see what 4.8 million calls actually look like.

The pandemic broke the baseline, and the city hasn't gotten quiet again.

The Data

I pulled the data through the Socrata API against the NYC 311 Service Requests endpoint, filtering on the noise complaint types (Noise - Residential, Noise - Street/Sidewalk, Noise - Commercial, Noise - Vehicle, etc.) - about 4.8 million rows from 2010 through October 2021.

The full 311 dataset is ~27 million rows since 2010, so noise is about 18% of everything New Yorkers call the city about. That alone is interesting. The Socrata API made the pull tractable - no need to download 25GB of CSV.

The Pandemic Rebase

Here's every month of noise complaints since 2010 - the 2010s in grey, 2020 onward in the accent:

Monthly 311 noise complaints, all seven noise types, January 2010 to October 2021. Regenerated in 2026 from NYC Open Data's archived and current 311 datasets - counts run somewhat higher than the original October 2021 pull because 311 backfills records.

For most of the decade, monthly noise complaints ran between roughly 15,000 and 45,000, swelling every summer (warmer weather → windows open → louder streets) and settling every winter. The busiest month of the entire pre-pandemic decade was June 2019, at about 59K.

Then 2020 happens. The line climbs from ~29K in February to over 100K in June - nearly double the loudest month of the prior decade - and the shape of the curve changes: it doesn't taper off in fall the way it usually does. Against the original October 2021 pull, 2020 logged 791K noise complaints to 2019's 471K - up by two-thirds in a single year.

The story doesn't end with the pandemic, though. 2021 stays elevated. Through the summer the city is running 60-90K complaints per month, roughly double the pre-2020 baseline, with no sign of returning to 2010s normal.

Where the Calls Come From

ZIP-level view of where complaints concentrate, 2010-2021:

Five ZIP codes account for a disproportionate share (counting every noise type, 2010 through October 2021):

The top spot - 10466 in the Bronx - logs more than half again as many complaints as the number-two ZIP. Most of the action concentrates in upper Manhattan and a couple of Bronx and Brooklyn ZIPs. Downtown Manhattan, Staten Island, and most of Queens stay relatively quiet on the 311 record.

That doesn't mean those areas are objectively quieter - only that fewer people there are calling 311. The two questions aren't the same. Reporting behavior matters too: how willing residents are to call the city, whether they expect anything to come of it, whether the noise lands on a weekend when they're home to hear it.

When the Calls Come In

If 2020 broke the baseline, what time of week did the breaking happen? Day-of-week noise volume in August, year by year:

Every year has the same shape - quiet Monday through Thursday, slow climb Friday, peak Saturday, taper Sunday. In a normal year, the Saturday peak is around 5,000-7,000 complaints for the month.

In August 2020, Saturdays hit ~20,000 - roughly 4× the prior decade's Saturday volume. The weekday baseline barely changed; the whole spike was concentrated on Fridays and Saturdays.

2021 cooled off relative to 2020, but Saturdays still ran ~12,000 - still double the pre-pandemic norm. Whatever happened in the summer of 2020 reshaped how loud New York's Saturdays get, and the pattern is sticky.

So Is New York Noisy?

By 311's definition, yeah - and more so than it used to be. The pre-pandemic numbers were already among the highest of any 311 category. 2020 didn't add a temporary surge; it shifted the floor.

Whether that reflects more actual noise or more people willing to call about it, I can't tell from this dataset alone. Probably some of both. More people working from home meant more ears for late-night activity. More outdoor bar service and street life in 2020 meant the activity itself shifted. And once a behavior - calling 311 about your neighbors at 1am - becomes normalized, it tends to stay.

Tools and Methodology

• Data: NYC 311 Service Requests, pulled via Socrata API. Filtered on seven noise complaint types (Residential, Street/Sidewalk, Vehicle, Commercial, Park, House of Worship, and plain "Noise"). 4,810,251 rows pulled, 4,797,737 after dropping rows with missing ZIP or borough.
• Spatial join: ZIP shapefile from NYC Open Data, merged on the cleaned complaint count.
• Tools: Python, pandas, NumPy, GeoPandas, fiona, matplotlib, seaborn, plotly.express. BeautifulSoup for some scraping. urllib for the API calls.
• Limit query: The Socrata API caps at 50K rows per call by default, so I bumped $limit to 300M and let it paginate.

Originally published October 26, 2021 as part of Introduction to Urban Data and Informatics, Columbia GSAPP (Professor Boyeong Hong). Notebook and source plots on GitHub. Earlier version on Medium. Monthly chart regenerated July 2026 from the NYC Open Data archived and current 311 datasets.