Select a data series on the right and it will display the appropriate chart with the data displayed beneath. A few helpful definitions. The data was analyzed two ways. FULL ZIP CODE compares Trump-branded condominiums to non-Trump condominiums in the same ZIP code. QUARTILE ONLY compares the two, but only uses the prices of non-Trump condominiums in the same price bracket as the Trump properties.
This is in turn analyzed by month, quarter or year, where appropriate. All rolls the data together in a single view across all properties, while similar views are made of properties in NYC (New York City), NYS (New York State) and FL (Florida).
Given the topic, it's important to understand how this data was collected, filtered and analyzed. Factba.se is committed to both sharing the sources of data, as well as the process and methodology behind it. If you have questions, they will almost certainly be answered here. It may be boring, but being informed isn't easy. It is, however, worthwhile.
To set the boundaries, the data was focused on residential condominiums bearing the Trump brand. As the exercise was to measure the Trump brand and its impact on the value of the properties, his ownership was not required. Apartment units, such as Trump Bay Street, were not included as they are primarily rental properties. In addition, The four Trump properties on Riverside Drive (120, 200, 220, 240) are in the process of a lawsuit to remove the Trump brand. As that could impact the study, these units in ZIP code 10069 were not included. Only one property, the Trump Park Residences in Shrub Oak / Yorktown, New York, was not included due to the paucity of data. Only four transactions were found in the public records. As such, it was not included.
This left the properties used in this analysis. Data was collected from nine sources, including: Broward County Property Appraiser, Connecticut Open Data, Monmouth County Tax Records (which include Hudson County), Palm Beach County Property Appraiser, public-record.com for Cook County, Gannett's lohud.com for Westchester, Miami-Dade's comparable sales tool and the incredible open data project at New York City's Department of Finance, which has a standard for open data every municipality should follow. We also included records from Zillow for recent sales. We gathered only condominimum sales within the 14 ZIP codes with Trump-branded properties.
Where the size of the unit was not present in the records, our AI, Margaret, was trained to search Google and, upon finding an exact match of the unit and address (with appropriate fuzz for Street vs. St, etc), verify square footage from at least two sources on a reserved list, such as Streeteasy or RedFin.
All transactions that were below market, referred to as not being at "arm's length", were removed. These are often transactions involving transfer between family members at a zero recorded value, or a foreclosure or other incident. Any property with clearly false data (a $250,000,000 sale) or missing or incorrect square footage data was similarly removed. This left 33,463 property sales of the 48,419 gathered to include in the analysis, or 69.1%
We first created two data cohorts per ZIP code. One included all condominimum units in that ZIP code not branded Trump, to compare against the units branded Trump in that ZIP code. For the units not branded Trump, we created a subset of only buildings in the same quartile (e.g. 0-25%, 25-50% and so forth) as the Trump properties in that ZIP code. The groupings are available in the downloadable dataset. This provides a closer comparison, but involves less data and can be more volatile.
We also created an index to compare the full dataset. This index established, for both the full ZIP dataset, and the Quartiles, a starting point of January, 2010 and a ratio between Trump-branded properties and non-Trump properties. The index then tracks that ratio on a monthly, quarterly and annual basis to measure its performance relative to its start in 2010.
For both of the above, this was done for each building, area, and the entire cohort on a monthly, quarterly and yearly basis. Due to the data density, a monthly index is done only for the full dataset. Quarterly datasets are indexed just for New York State and New York City. All other data slices do not have enough transactions for the index to be an effective measure. A comparison of average sale price per square foot is used in these cases.
How we assembled this data and analyzed 48k+ records. Hint: robots were involved. And caffeine. And lots of screaming and cursing at the data Gods...Factbl.og
Keep up with the latest from Factba.se
@FactbaseFeed. Boring by choice.
Specs and Sale Price Per Square Foot through the last calendar year.