In total, 398 distinct estates were plotted; the following observations were made by examining the maps created from these points. By turning on and off different layers, looking at data sets in tandem with each other, and comparing data between current and historic maps, I noticed several patterns.
General observations
The first general point I noticed is that Nassau County, the more western of Long Island’s two counties, was a significantly more popular location to build a home. This might be because it is closer to New York City than Suffolk County is, meaning the travel time between the two locations was shorter. Proximity would be important for people who might have made their full-time home on the island, especially if they were commuting into the city regularly.
Within Suffolk County, it is clear that the preferred spots were as close to the county line with Nassau as possible, with several houses being built just east of the divide. Houses within Suffolk are also concentrated in the western part of the county; the house the furthest east is Carefree Court in Asharoken. This means that all of the houses in Suffolk County are clustered within approximately the closest six miles to the county line, while the other 80 miles did not have Gold Coast estates.
This is not to say houses were not built further east; just that no Gold Coast houses were. One explanation for this trend is that people looking to build a Gold Coast estate did not want to be that far east. However, it is also possible that later historians decided where the eastern edge of the “Gold Coast” was, and thus the data I had available only listed houses built within that geographic range.
For both counties, I noticed that many estates were grouped along the coast. There were also many houses built further inland, but on Great Neck and Locust Valley in particular, the coastline is filled with estates. Much as today, these property locations were most likely sought after for their views and water access, so this pattern is not particularly surprising, but still worth noting.
Acreage
The first, and most frustrating, point regarding acreage is that the biggest “category” is unknown. Of the 398 estates that were mapped, 142 of them do not have acreage marked, making it difficult to draw any decisive conclusions. Additionally, the acreage that does exist for many estates was recorded in the late 1920s, the 1930s, or the 1940s – potentially fifty or more years after an estate was built. Because of this, the reliability of this category is difficult to ascertain.
With that being said, the largest category for houses with an actual acreage number were those with 11-20 acres (49 estates), followed by those with 1-10 (37 estates), 101-200 (30 estates), and 21-30 (24 estates). There are two main conclusions that might be drawn from the surplus of estates with low acreages: either the size of estates had significantly shrunk by the time its acreage was recorded, or many people were building houses located on smaller parcels of land. If the latter conclusion is true, this suggests that perhaps the larger estates were built by the wealthiest people of this era, while the smaller ones were built by the less (but still comparatively) wealthy. Since there are more “normal wealthy” people than “mega-wealthy” people, it makes sense that there would a greater number of smaller estates.
I also noted that the estate acreage across the area as a whole didn’t seem to follow much of a pattern. Huge estates were next door to small estates as well as medium estates. Along the coastline of Great Neck estates seemed to stay a bit smaller, but this was the only spot I noticed distinct patterning in.
I also noticed that the two 1,000+ acre estates, Caumsett and Jericho Farms, are both still standing today as a park and country club respectively. It’s difficult to tell when looking at the map as a whole if larger estates were more likely to survive, but it was worth noting that the houses built on the two largest pieces of land are still standing today.
Year of Construction
In terms of the popularity of construction periods, the most popular timeframe was 1910-1919 (142 estates), followed by 1920-1930 (130 estates), 1900-1909 (91 estates), and 1890-1899 (34 estates). The low numbers at the beginning of my time frame suggest that the construction of estates started slow, reached its heyday during the period of 1910-1919, then dipped slightly from 1920-1930. The high construction numbers later in the date range make sense since more people had earned enough money by then to afford a mansion, but also shows that the popularity of these mansions did not wane with time. It was most likely factors outside of individual control, such as the Great Depression and World War II, that caused the end of the Gold Coast Era as opposed to a waning interest in estate creation.
The years of construction tend to be dispersed geographically, although early estates were generally built either on the coastline, or near major roads at the southern end of the geographic area. This is reinforced in the isolated map, which shows the coastlines being filled first, followed by the main roads, and finally the “in between” areas. As previously discussed, the coastline would have been an attractive spot for many people. For those looking for easy transportation back to New York City, being near major roads would also be advantageous. I also noticed that many of the houses built from 1920-1930 are clustered around the area of Glen Head and East Norwich, perhaps suggesting that this inland area was the best land available by the time they were built.
Architect
Architects were by far the most diverse category mapped, with 151 different architects, not including the “missing” category. Of those 151 architects, 15 designed five or more houses, and 55 designed at least two. This leaves almost 100 architects who only designed one house. Because of these numbers, it was impossible to map each individual architect, and thus only the ten with the greatest number of houses appear on the map, with everyone else grouped into “Other.”
One trend that I noticed was that of the fourteen estates that were financed by an architect, thirteen of them were designed by that architect as well. Augustus Cass Canfield, a naval architect, is the only architect who didn’t design his own home. This made me wonder if some architects in this area were able to make a significant amount of money designing Gold Coast estates, and then in turn build their own Gold Coast estate.
I found that several architects seemed to prefer geographic areas; for example, the two houses designed by Harry Ellingwood Donnell were both located in Eatons Neck, and seven of the ten houses designed by Charles Pierrepont Henry Gilbert are also clustered.
Delano and Aldrich were by far the most prolific architects with twenty estates; Charles A. Platt and Peabody, Wilson and Brown were the second most popular with eleven each. However, only two of Delano and Aldrich’s estates are located on the water, and the vast majority are at the southern bound of the geographic area. This made me wonder if perhaps this architecture firm was less prestigious than others, and therefore, people who couldn’t afford coastline property but could afford an estate were able to hire them.
Finally, several architects seemed to prefer certain architectural styles over others. Seven of Charles A. Platt’s eleven houses were Georgian Revivals, and two of Horace Trumbauer’s five were Federal. The two Federal houses designed by Trumbauer are also very close to each other and were built several years apart – making me wonder if the second financier liked the first’s house so much they wanted a copy of it.
Architectural Styles
Similarly to architects, architectural styles were also a difficult category to make sense of. Eighty different styles were noted, not including the missing ones, but there were also a number of similar sounding styles. For example, there were eight styles with the word “Italian” in the name: Italianate, Italian Farmhouse, Italian Neo-Classical, Italian Renaissance, Italian Renaissance Villa, and Italian Villa. Since I do not know architecture well and the style of a house can be subjective I did not feel comfortable trying to group “similar” styles together, thus this data is a little hard to make sense of.
One pattern that does emerge is that more stately styles seemed to be built often, with Georgian Revival (66 estates), Federal (43 estates), Tudor (43 estates), and Colonial Revival (33 estates) being the most popular. Revival styles in general were also popular, with a number of variations such as Elizabethan Revival, Federal Revival, and French Classical Revival being included. Nine of the estates were “20th Century Eclectic” – which for the time would have been more modern than many of the other styles. This raises the possibility that the people financing these houses, as well as other houses with “eclectic” styles, were from the new money class and wanted a newer style house to reflect that.
I also noticed that many estates built between 1910-1919 were Georgian Revival. Since that was the most popular style as well as the most popular decade of construction, however, it’s hard to say if it’s because of those factors or if Georgian Revival was a genuinely popular style during that period.
In creating the isolation map for architecture style, I noticed that Tudor houses seemed to be primarily grouped in Oyster Bay, with approximately 25 of the 43 Tudor houses located there. Additionally, Colonial Revivals seemed especially popular along the coast; there is also a small cluster of five Colonial Revival houses grouped together in Great Neck. As discussed in the Historic Maps section of this analysis, this same location was home to the North Country Colony; suggesting that the architectural similarities were a purposeful decision to create unity among the homes of this area.
Financier
Interestingly, there were several financiers who built multiple houses during the period of 1890-1930. For some this was because they sold their first house and built a new one, but for others it seemed that they lost their original house to fire or another destruction, and then rebuilt in a similar location. George Ernest Fahys Sr. for example, built two houses called Hillaire near Locust Valley, although the first Hillaire was still standing when he built the second. George Bullock built two houses named Folly on the same location after the first one burned down; he later built another house up the road called Yeadon. In perhaps the worst luck of all, Paul Drennan Cravath built three houses named Veraton; the first two burned down, but Vertaon II and III were at the same location. He later built a fourth estate called Still House near the location of Veraton II and III.
Some families also seemed to group their estates nearby each other. The Guggenheims for instance had three separate estates near Sands Point, while the Pratts had seven estates built closely together.
Industry of Financier
The first notable point about industries is that finance was by far the most popular. 101 estates were built by people working in finance; the next largest group was 55 people who worked in law. This trend makes sense when you consider the prominence Wall Street gained during this era, as well as the fact that people could easily become wealthy working in finance during the Gilded Age.
Additionally, people who worked on Wall Street were most likely not the mega-rich heads of industry. This means that there would be more people working in finance, and that this group might have considered their estate a full-time home rather than a seasonal escape if they still had to go to work regularly. The prevalence of financiers then reinforces my theory that many of the smaller houses might have been built by less wealthy people looking for a full-time home, contributing to both the number of estates and their locations near major transit routes. These comparatively less wealthy people also might have wanted to build homes in order to give their newfound wealth legitimacy, or as a status symbol.
After creating an isolated map for the industry category, it was a bit easier to see potential patterns. Generally, people were still scattered, but there were several small groupings across industries. Lawyers seemed to prefer coastlines or main roads, a couple of people who worked in media were nearby to others, and there were three pairs of grouped houses for those in utilities. But despite this data isolation, there still aren’t clusters of industries the way one might expect.
Similar to many other groupings the industry category was varied. There were 36 distinct industries, twenty of which had at least five people working in it. Patterns in industry are a bit difficult to see, however, since everything is a bit overwhelmed by the finance category. The financiers appear to be grouped, but that might just be due to the sheer number of them rather than a purposeful attempt to build near their peers. There were also many houses built in the Georgian Revival style by financiers, but given both the popularity of the profession as well as the style, it’s hard to say if this is a distinct trend.
Year Demolished
The first surprising point of information in this category is that most demolished houses were destroyed during the period of 1950-1959. Especially given the prominence of people working in finance I had thought many houses would have been demolished in the 1930s due to the Great Depression, but only eight were torn down during that time. By comparison, twenty-nine houses were demolished from 1950-1959, and twenty-two from 1960-1969. This suggests that the economic crisis of the 1930s might not have impacted the homes the way one might think, and rather it was the development of the island post-World War II and a rise in property tax rates that were larger contributing factors.
It is possible that even if estate owners did not have the money to keep their house immediately following the Depression, there was also no one wealthy enough during this time to redevelop the land. Additionally, there was very little need for housing developments on Long Island prior to suburbanization efforts in the 1950s, meaning that houses might have been foreclosed on or abandoned earlier, but the actual removal of the estate did not happen until developers had the money and incentive to perform this work. The development of highways and major roads also did not occur until later, so houses would not have been destroyed in the 1930s to make way for those structures.
Only thirteen houses total were torn down prior to 1940, and many of them were rebuilt soon afterwards. These removals may have been accidental, such as a fire destroying the home, rather than a purposeful demolition. The number of houses being razed also slowed down significantly after 1970-1979, but there are still several demolished in each period, including one in 2020.
Geographically, about half the demolitions that occurred during the period of 1960-1969 were in Great Neck, indicating specific development of the area during that time. Looking at the isolation map, thirteen of the twenty-two demolitions during this same period were for houses on the water, also speaking to redevelopment. Additionally, six of the eleven houses torn down between 1970-1979 were located near the Long Island Expressway in Old Westbury, suggesting that highway construction during this period might have been the impetus for those houses’ removal.
Current Use
One heartening piece of information about the current state of Gold Coast homes is that the majority are still standing. Of the 398 houses that were mapped, the current status is only known for 349 of them. 139 of those have been demolished (~40%), while 210 are still standing (~60%). 147 are private residences (~42%), with the next biggest category of use being educational buildings.
Many of the private residences are clustered near Oyster Bay, while the coastline, particularly in Great Neck, has mostly been demolished. The coastline demolition might be because those houses degraded quicker due to environmental damage caused by being near the ocean, but it might also be due to the fact that land prices are more expensive and thus, more attractive for developers. The clustering of private residences is interesting, and my best guess for its meaning would be that the values of the community and people who settle there might be different. There might be a desire to preserve the old rather than tear it down for the new, or there could other factors that prevented developers from buying land in Oyster Bay. A lot of demolition also occurred near the Long Island Expressway, which is not surprising.
Many of the homes that were in Suffolk County have been demolished. Only 42 Gold Coast estates were in that county originally, and of those, sixteen have been demolished, eleven are private residences, and seven are of unknown status. The remaining eight houses have a myriad of uses.
I did not notice any factors that correlated with the current use of a house other than demolition and private residence – those that are serving new purposes are scattered, and don’t seem to have any unifying details.
Historic Maps
Comparing the historic maps with various data overlays, the first obvious note is that the island becomes increasingly more developed as time progresses. The earliest map from 1906 shows a fairly undeveloped area with only a few main roads and large parcels of land. By the 1944 maps, however, the land parcels have shrunk, the number of roads have increased, and clusters of towns with clearly defined centers have begun to emerge. This points more to the development of the island than anything to do with the estates specifically but could help explain why so many were demolished in later years.
I also noticed that many main roads, such as Jericho Turnpike, appeared around the same time the houses did. This could point to either the roads being built so that people could more easily reach their homes, or people purposefully building their homes near main roads. Without further research, however, it’s hard to say for sure which came first.
In the 1922 map by Baker Crowell Inc, it’s incredibly glaring that the coastlines filled up before more inland houses did. Looking at the size of the land parcels as well as the construction years, one can see that the coastal areas had much smaller land parcels than more inland areas by 1922, and that many of the later estates were built further from the ocean. This could suggest both that people built inland because the coast was already taken, as well as that one could most likely buy a larger amount of land for comparably less if they went further inland.
One area on the 1923 E. Belcher and Hyde map already begins to resemble a development and was completely built upon by 1919. The “North Country Colony” already had defined roads, smaller land parcels, and a significant number of houses by 1923; yet almost half of these homes have since been demolished.
According to the 1929 Dolph and Stewart map, most of the houses built in Suffolk County seemed to have been built away from main roads. Similarly, the 1929 map also shows many houses keeping away from main roads, although it is not known if this was done on purpose to accommodate the people living in these estates, or if the houses predated the roads so they were built further away.
Next page: Conclusions