A truck parked in a metered or residential space is supposed to have a temporary no-parking permit for it. Those permits are how I know to warn someone before a truck permit hits their block, and I check both cities for new ones every day. For this report I pulled every moving-truck permit on file for August 21 through September 6 and counted them by neighborhood and by block.
Neither city tells me whether a truck is loading or unloading, so I inferred this from the permit dates. A permit that ends on or before August 31 belongs to a tenant moving out. A permit that starts on September 1 or later belongs to a tenant moving in.
Boston filed 1,204 permits for September 1 and Cambridge filed 226. Boston averaged 137 a day over August 21 to 24, so September 1 runs 8.8 times higher than the weekend before it. The volume drops back within three days. I count only 21 permits between the two cities for September 6.
Moving out, moving in
Boston and Cambridge between them see 2,650 permits ending between August 25 and 31, and 1,763 starting on September 1 or 2. That works out to 1.50 departures for every arrival. The two cities land on almost the same figure, 1.50 and 1.52.
Moving out, Aug 25 to 31
2,650
2,201 in Boston and 449 in Cambridge, across seven days
Moving in, Sep 1 to 2
1,763
1,468 in Boston and 295 in Cambridge, across two days
Ratio
1.50 : 1
permits ending in the departure week for every one starting on Sep 1 or 2
The two windows on either side of September 1 are deliberately not the same length. Departures spread across a full week because a tenant can book the truck for any day the lease allows. Arrivals compress into 48 hours because nobody takes possession of their new apartment before the lease starts. The counts therefore describe how many permits fall in each period. They do not measure a net gain or loss of residents, and the Net column in the table below represents the difference between the moving-out and moving-in counts on either side of September 1.
Boston neighborhoods
Permits out against permits in
Ending Aug 25 to 31 against starting Sep 1 to 2
Out
In
818 permits were filed for South Boston over the nine days, more than for the next two neighborhoods combined. Allston-Brighton drew 639, the North End 395, Beacon Hill 297 and Back Bay 263. Not one moving permit was filed in Mattapan.
Charlestown and Jamaica Plain lean hardest toward departure, at 2.8 and 2.6 permits ending in late August for every one starting on September 1. South Boston and Allston-Brighton run closer to 1.2, so most of what leaves those two neighborhoods is replaced within two days.
9.3%
of Boston moving permits are on a block that is scheduled for street cleaning the same day. 435 of 4,677.
15.5%
of Boston moving permits share a block with a construction permit that is active the same day. 723 of 4,677.
Cambridge neighborhoods
Permits out against permits in
Ending Aug 25 to 31 against starting Sep 1 to 2
Out
In
795 permits were filed across Cambridge over the same nine days, about a fifth of Boston's total, and 226 of them fall on September 1. Mid-Cambridge leads at 147. The Port and Wellington-Harrington tie at 99, followed by East Cambridge at 94 and Cambridgeport at 84.
The Port comes closest to even, at 51 permits out against 41 in. Cambridgeport and Mid-Cambridge lean furthest toward departure, both near 1.8 to 1. Not one moving permit was filed in Area 2/MIT, which is mostly institutional dormitories.
8.2%
of Cambridge moving permits are on a block that is scheduled for street cleaning the same day. 91 of 1,111.
6.3%
of Cambridge moving permits share a block with a construction permit that is active the same day. 70 of 1,111.
Density: permits per mile
South Boston has 74 miles of street for its 818 permits. The North End has 13.8 miles for 395, and Beacon Hill has 10.2 miles for 297. When I divide permits by street miles, the whole list reorders.
Beacon Hill reaches 29.1 permits per mile and the North End reaches 28.6. Back Bay comes third at 15.6. The two leaders carry roughly twice the density of anywhere else in either city, and about sixty times the density of Dorchester.
Permits per mile of street
Nine-day total divided by centerline miles
Boston
Cambridge
The busiest blocks
Sheafe Street in the North End runs one block, between Margaret Street and Snow Hill Street. I found 24 permits on it across the nine days, more than on any other block in either city. Saint Germain Street between Dalton Street and Massachusetts Avenue holds 20. Harvard Street between Hancock Street and Dana Street in Mid-Cambridge holds 19, the most of any block outside Boston.
Counted by street rather than by block, Commonwealth Avenue in Allston-Brighton leads Boston at 141 permits, more than double Beacon Street in Back Bay at 68. Marlborough Street follows at 58, then Hanover Street in the North End at 45. Cambridge's busiest streets are Harvard Street at 28 and Columbia Street at 24.
On September 1 alone, 775 Boston blocks and 176 Cambridge blocks hold at least one truck permit. Across the full nine days those totals reach 1,511 blocks in Boston and 419 in Cambridge, which is one block in twelve in Boston and one in six in Cambridge.
When moving trucks aren't the only thing you need to worry about
A moving truck rarely has the block to itself. I checked every moving permit against the sweeping schedule for the same block on the same day, and against every construction permit active on that block that day.
Sweeping clashes on Sep 1
169
124 in Boston and 45 in Cambridge, out of 526 across the whole period
Construction clashes on Sep 1
195
177 in Boston and 18 in Cambridge, out of 793 across the whole period
All three on Sep 1
6
blocks carrying a truck, a scheduled sweeper and live construction at once
The sweeping clash breaks down as 435 permits in Boston and 91 in Cambridge. On September 1 alone there are 169 of them, 124 in Boston and 45 in Cambridge. Anyone parking on one of those blocks that morning is competing with moving trucks and the street sweeper, and with the tow truck that precedes it.
The construction clash is larger and lands almost entirely in Boston, 723 permits against 70 in Cambridge, because Boston runs far more occupancy work. September 1 carries 195 of them. Construction permits do not pause for moving day, so one block can lose a stretch of curb to a dumpster and another stretch to a truck on the same morning.
Sheafe Street in the North End is the heaviest double. Seven of its 24 permits fall on September 1, and the block sits on the sweeping schedule that day. Hanover Street between Cross Street and Mechanic Street carries five under the same conditions, and Columbia Street between Worcester Street and Harvard Street in The Port carries four.
Six blocks draw the full trifecta on September 1. Brackett Street in Brighton, between Faneuil Street and Washington Street, is the worst of them: three moving permits, a sweeper due that day, and construction already holding part of the curb. The other five are Prince Street and Commercial Street in the North End, Lexington Street in East Boston, Vinton Street in South Boston, and Harvard Street in The Port, each with a single truck. May the odds be ever in their favor.
Every neighborhood
| Neighborhood | Out | In | Net | Sep 1 | Peak 9d | Miles | Per mile | Out : In |
|---|
How I counted this
- Sources. Boston figures come from the City of Boston street occupancy permit feed, filtered to moving-truck permits. Cambridge figures come from the City of Cambridge permit feed. I pulled both on August 28, 2026. Somerville, Chelsea, Malden and Everett publish no comparable moving-permit feed, so I left them out.
- Direction is inferred. Neither feed tells me whether a truck is loading or unloading. I treat a permit ending on or before August 31 as a departure and one starting September 1 or later as an arrival. That reading fits a lease-driven calendar, but it is not a fact about any individual permit, and some permits are neither.
- The two windows cover different amounts of time, seven days out against 48 hours in, because that is how the two events actually distribute. Out, In and Net are therefore raw permit counts and not a like-for-like comparison of flow. Comparisons between neighborhoods use identical windows on both sides, so those hold up.
- Permits are not trucks. One move can pull two permits, for both sides of a street or for two adjacent blocks, and plenty of moves pull none at all. Every count here is a floor on activity rather than a census of vehicles.
- The feed is bounded at both ends. Looking back, both feeds keep only about a week of expired permits, so I cannot compare September 1 against an ordinary Tuesday in June. The 8.8 times figure compares it to the daily average of August 21 to 24, a stretch already carrying early movers, and against a quiet week the multiple would be higher. Looking forward, Boston publishes 13 days ahead and Cambridge a median of 11, so nothing after September 10 is visible to me today, and the last few days inside that window are still filling in.
- Sweeping clashes are computed, not read. I evaluate each block’s stored sweeping rule against the permit date, matching day of week, week of month and season the same way the app does when it schedules a reminder. A clash means the sweeper is scheduled for that block that day. It does not mean the sweeper arrived, and it does not account for a holiday suspension. Construction clashes are a plain date-range overlap on the same block, counting construction and other occupancy permits together.
- Permits carry dates, not times. Every permit in both feeds starts at midnight local time, so there is no time-of-day signal for me to analyze.
- Neighborhood assignment comes from the block each permit snaps to in my street data rather than from the permit's own address text. I merged overlapping labels in the source data: Allston/Brighton and Brighton became Allston-Brighton, and the three Dorchester labels became one. Five Boston permits landed on blocks with no neighborhood assigned. Those five sit in the city totals but not in the neighborhood table, which is why the columns fall five short.
- Miles are geodetic centerline miles of the active street segments on record for each neighborhood.
It looks like more people moving out than in. Are people leaving Greater Boston?
I cannot tell yet.
The raw figures look like an exodus. The two cities record 2,650 permits ending in the departure week against 1,763 starting on September 1 and 2, which is 1.50 to 1. Most of that gap is an artifact of the windows, seven days of departures set against two days of arrivals.
Matching the windows removes most of it. Departures ending August 25 to 31 come to 2,650. Arrivals starting September 1 to 7, the same seven days on the other side of the lease date, come to 2,204. That is 1.20 to 1.
The rest shrinks once the forward limit of the feed is taken into account. Boston publishes only 13 days ahead, so arrivals late in that week are still being filed. Correcting for the rate at which Boston permits arrive late brings the figure to roughly 1.1 to 1. That last step is an estimate rather than a count, so treat it as a direction and not a number.
A ten percent excess of departures is not evidence of an exodus. Someone moving from one Boston apartment to another pulls a permit at each end and lands on both sides of the ledger. Someone leaving the region counts only as a departure and someone arriving from outside counts only as an arrival, so a genuine net outflow would show up here. So would several quieter explanations, including a simple difference in how willing people are to file a permit when they leave compared with when they arrive.
One year of data cannot separate those. This sample leans toward more people leaving than arriving, and it leans far too gently to say so out loud. I would need the same nine days in 2027 and 2028, measured the same way against the same feeds, before the direction of the gap carries any weight. I plan to run this again next year.