SW10 digs deep

By Carl Goodwin

(This article was first published on R – thinkr, and kindly contributed to R-bloggers)

Responding to a weak property market

In December I looked at how recent events have shaped the property market in London SW10. If short-distance moves are off the table in the current climate, how are property owners responding? When sales are weak, are planning applications in the ascendency? I applied data science techniques to Royal Borough of Kensington and Chelsea (RBKC) planning data to find out.

Property transactions evaporated with the Financial Crisis. The Government “stamped” on the green shoots of recovery with penalising duties on moves. And the uncertainty surrounding Brexit hasn’t helped.

Property development offers an alternative way to add space. Owners unable to sell would want to consider their options, engage consultants, and secure planning permission. So one could reasonably expect the data to reflect a delayed response. And that’s what we see when plotting sales versus planning applications.

From 2001 to 2017 in SW10 there were circa 6,800 property transactions, and 6,400 decided planning applications. The numbers are surprisingly balanced in both their totals and their shape over time. When sales are down, attention does appear to shift to development.

Is adding space the motivation?

So owners are responding to the weaker market by doing something to their properties. But what? Given the state of the market, is adding space a key theme?

Planning applications, at least on the RBKC planning search portal, are not naturally categorised by type of change. So, as a starting point, I used a “bag of words” data science technique to explore the application proposal descriptions for word frequency.

I stripped out stop words, converted proposal descriptions to lower case, and stemmed the remainder. This ensured words like “Windows” and “windows”, as well as “terrace” and “terracing”, were treated as one.

The resultant word frequency plot includes the “what”, for example, “extension”, as well as the “where” and “how”. From this I could pick out a reasonable set of broad “what” themes, such as extensions and basements. If, for example, a proposal for an extension also mentioned windows, then I counted this once only as the bigger theme.

When sales are down, dig down

With a reasonable set of themes, I could now visualise how the number of planning applications has changed over time by type of development. Adding space, down, sideways or up, does appear to have driven much of the uptick.

Finding space though in dense and tightly-regulated conservation areas is challenging. So it is perhaps unsurprising that “down” is sometimes the only, albeit more expensive, way to go.

The majority of applications, across all themes, are decided favourably. Often, the planning review process will drive the need for the applicant to make proposal changes, or submit additional information. In such cases, the applicant withdraws the proposal, updates, and goes again.

By applying a linear regression model to each of the themes, we can better see which types of planning applications have grown more steeply in recent years.

The most significant change has been in the volume of applications related to trees. However, it is worth noting, in this particular case, that RBKC adopted a Trees and Development Supplementary Planning Document on the 20 April 2010.

Steepest trends by planning application theme

We can more clearly see that adding space is a key motivation for the growth in planning applications against the backdrop of the weaker property market. In contrast, it does seem reasonable that owners undertake roofing works more out of necessity. So this type of development might be a little more immune to the headwinds and tailwinds of the property market.

Square-footage is king

An entirely separate and interesting topic of exploration might be the net economic effect of the shift from sales to development. One could anticipate negative effects on one ecosystem, and positive effects on another. For example, bad for estate agents and removal firms, but good for builders and planning consultants.

But to return to the opening question, in response to a weak property market, it would seem that SW10 is digging deep, both into its pocket and the ground, to find more living space. “Square-footage is king”, as central London estate agents like to say, so surely it’s a good long-term investment?

R toolkit

Packages Functions
purrr map_df
rvest read_html; html_nodes; html_text; html_table; html_attr
SPARQL SPARQL
dplyr bind_rows; filter; mutate; mutate_at; if_else; select; group_by; summarise; arrange
tidyr spread; nest; unnest
broom tidy
stringr str_c; str_detect
lubridate as_datetime; date; year; dmy
tibble data_frame
tm removeWords; stemDocument
qdap freq_terms
ggplot2 geom_bar; geom_area; facet_wrap; geom_col; coord_flip
ggthemes theme_economist

The code may be viewed here.

Citations / Attributions

R Development Core Team (2008). R: A language and environment for
statistical computing. R Foundation for Statistical Computing,
Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org.

Planning application data web-scraped from the Planning search pages with the kind permission of the Royal Borough of Kensington and Chelsea.

The post SW10 digs deep appeared first on thinkr.

To leave a comment for the author, please follow the link and comment on their blog: R – thinkr.

R-bloggers.com offers daily e-mail updates about R news and tutorials on topics such as: Data science, Big Data, R jobs, visualization (ggplot2, Boxplots, maps, animation), programming (RStudio, Sweave, LaTeX, SQL, Eclipse, git, hadoop, Web Scraping) statistics (regression, PCA, time series, trading) and more…

Source:: R News

Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.