The New Urban Success: How Culture Pays
227The New Urban Success: How Culture Pays
DESISLAVA HRISTOVA,
Cambridge University, Cambridge, UK
LUCA MARIA AIELLO,
Nokia Bell Labs, Cambridge, UK
DANIELE QUERCIA,
Nokia Bell Labs, Cambridge, UKUrban economists have put forward the idea that cities that are culturally interesting tend to attract “thecreative class” and, as a result, end up being economically successful. Yet it is still unclear how economic andcultural dynamics mutually influence each other. By contrast, that has been extensively studied in the caseof individuals. Over decades, the French sociologist Pierre Bourdieu showed that people’s success and theirpositions in society mainly depend on how much they can spend (their economic capital) and what theirinterests are (their cultural capital). For the first time, we adapt Bourdieu’s framework to the city context. Weoperationalize a neighborhood’s cultural capital in terms of the cultural interests that pictures geo-referencedin the neighborhood tend to express. This is made possible by the mining of what users of the photo-sharingsite of Flickr have posted in the cities of London and New York over 5 years. In so doing, we are able toshow that economic capital alone does not explain urban development. The combination of cultural capitaland economic capital, instead, is more indicative of neighborhood growth in terms of house prices andimprovements of socio-economic conditions. Culture pays, but only up to a point as it comes with one of themost vexing urban challenges: that of gentrification.Additional Key Words and Phrases: culture, cultural capital, Pierre Bourdieu, hysteresis effect, Flickr
Original paper published on Frontiers: https://doi.org/10.3389/fphy.2018.00027
The French sociologist Pierre Bourdieu argued that we all possess certain forms of social capital. Aperson has, for example, symbolic capital (markers of prestige) and cultural capital (knowledge andcultural interests). These are forms of wealth that individuals bring to the “social marketplace”. Hiswork ultimately had the goal of testing what he called ‘the differential association’ hypothesis [3, 11].This states that individuals with similar composition of capital are more likely to meet, interact,form relationships, have similar lifestyles and, as a result, be of the same social class. In his surveysof French taste, Bourdieu proved this to be the case. In so doing, he also found what he called the hysteresis effect , which refers to any societal change that provides opportunities for the alreadysuccessful to succeed further. During times of change, individuals with more economic and culturalcapital are the first to head to new (more advantageous) positions. A similar argument could applyto cities as well: a city constantly changes, and neighborhoods with more economic and culturalcapital will be the first to head to new positions, contributing to the city’s economic success.Such an argument has not been widely studied in the city context, yet it is behind most modernurban renewal initiatives inspired by the ‘creative class’ theory. This theory holds that cities withhigh concentrations of the creative class (e.g., technology workers, artists, musicians) show higherlevels of economic development [7]. The creative city as a planning paradigm supports creativityand culture by design, providing a direct link between cultural amenities, the quality of life, andeconomic development [10, 14, 24]. However, cities and neighborhoods which are consideredexemplars of creativity today are yet ridden with social and economic inequality [4]. Cities suchas San Francisco, New York, and London display a glaring gap between high- and low-incomeresidents [6]. It is therefore interesting to explore the complex interplay between economic success
Authors’ addresses: Desislava Hristova, Cambridge University, Cambridge, Cambridge, UK, [email protected];Luca Maria Aiello, Nokia Bell Labs, Cambridge, Cambridge, UK, [email protected]; Daniele Quercia, Nokia Bell Labs,Cambridge, Cambridge, UK, [email protected]. , Vol. 1, No. 1, Article 27. Publication date: April 2018. a r X i v : . [ c s . C Y ] A p r nd cultural creativity. The challenge is that it is hard to capture culture—all the more so at thescale of entire cities. We partly tackle that challenge by making two main contributions: • We quantify neighborhood cultural capital from the pictures taken in both London andNew York City over the course of five years. To this end, we build the first ‘urban culture’taxonomy which contains words related to cultural activities and groups these words intonine categories. We create this taxonomy by proposing a semi-automated 5-step approachthat uses both a top-down classification of the creative industries and a bottom-up crowd-sourced knowledge discovery from both Wikipedia and Flickr. We then select picture tagsthat match these words. These tags come from approximately 10M geo-referenced picturesin London and New York which were posted on Flickr from 2007 to 2015. As a result, eachneighborhood in the two cities is characterized by the fraction of picture tags that belong toeach of the nine cultural categories. • For the first time, we test Bourdieu’s hysteresis effect in the city context. We find that urbandevelopment is well explained by a combination of cultural and economic capital. Thiscombination allow us to successfully predict property values in New York and Londonneighborhoods (with R = .
56 and 0 .
81, respectively).
In the urban setting, culture is mainly produced through the cultural services and artifacts of thecreative industries. In 2015, the UK Department of Culture, Media and Sports adopted one of themost robust definitions of creative industries [2]. This includes nine macro-industries: • (100) Advertising and marketing • (200) Architecture • (300) Crafts • (400) Design: product, graphic and fashion • (500) Film, TV, video, radio and photography • (600) IT software and computer services • (700) Publishing • (800) Museums, galleries and libraries • (900) Music, performing and visual artsWe then needed to expand this coarse-grained categorization into a multi-level taxonomy whosetop nodes were these nine categories, intermediate nodes were subcategories, and leaf nodes wereterms related to culture. Both subcategories and terms were to be determined, and we did so in fivesteps, which are described next (Figure 1). Step 1: Wikipedia Mapping.
A way to iteratively expand the initial nine categories is to connectthem to an existing knowledge base of linked concepts. Because of its well-structured and hierar-chically organized content, Wikipedia was fit for purpose, so much so that it had often been usedto build semantically related large-scale taxonomies [19]. However, each of the nine categorieswas hard to map to a single Wikipedia page. To ease the mapping, we disaggregated the ninecategories based on their Wikipedia definitions. For example, we split ‘(400) Design’ into each ofthe elements defined in its description: ‘Product Design’, ‘Graphic Design’ and ‘Fashion Design’.After doing so for all the nine categories, we obtained twenty five Wikipedia (Article) categories:(101) Advertising; (102) Marketing ; (200) Architecture; (300) Crafts; (400) Design; (401) Productdesign; (402) Graphic design; (403) Fashion; (501) Film; (502) Television; (503) Video; (504) Radio;(505) Photography; (601) Technology; (602) Gaming; (603) Software; (700) Publishing; (801) Arts; tep1 Wikipedia mapping [n=25]
Maps creative industriesto Wikipedia pages
Step2
Wikipedia tree
Traverse the Wikipedia articletree to find related concepts [n=657]
Step3
Flickr graph expansion
Collect Flickr tags co-occurringwith the Wikipedia terms [n=29,495]
Step4
Wordnet similarity [n=633]
Filter out words that aresemantically unrelated to culture
Step5
Cleaning and validation [n=263]
Remove terms whose meaning isnot reflected in Flickr pictures
Finaltaxonomy"Is-a"Taxonomyfiltering n=441 n=533
Fig. 1. The five steps performed to obtain our taxonomy of cultural terms (802) Culture; (803) Museums; (804) Libraries; (901) Music; (902) Performance art; (903) Theatre;(904) Visual arts. We call these the top-level categories.
Step 2: Wikipedia Article Tree.
To expand these top-level categories, we use the Wikipedia graph.In general, Wikipedia category structure is essentially a graph of pages that can be navigated tofind concepts that are related to each other. Starting from the twenty five top-level categories, wecollected all the pages that directly link to them (that is, those that are 1-hop distance apart inthe graph ). After automatically removing community pages (which are not actual articles ) andmanually removing pages corresponding to highly ambiguous terms such as ‘color’, we were leftwith 657subcategories (connected to the initial 25 top-level categories). First ‘is-a’ filtering.
Not all the 657 subcategories are relevant. Our goal was to build a taxon-omy. By definition, a taxonomy connects categories and subcategories that are related with ‘is-a’relationships (e.g., the subcategory‘film’ is-a ‘product’). The ‘is-a’ relationships relevant to culturehad been identified by Gunnar Tornqvist in his book “The Geography of Creativity” [23] and wererepresented by what he called the 4- P s: { process , place , person , product } . Therefore, out of the 657subcategories, we filtered away those that were not { process , place , person , product } and kept theremaining 441 subcategories (e.g., Architect and Buildings are subcategories of Architecture).These subcategories form the second level of our taxonomy. Step 3: Flickr Graph Expansion.
To expand the taxonomy coverage as much as possible, weextended it with a third level containing specific terms related to the 441 subcategories. To do itwith a data source complementary to Wikipedia, we relied on exploiting the structure behind tag We do not navigate the graph at further hops because the number of connected pages grows exponentially at each hop,quickly including concepts that are highly unrelated. Community pages are not Wikipedia articles. Instead, they belong to the following Wikipedia categories: wikipedia,wikiprojects, lists, mediawiki, template, user, portal, categories, articles, images l l l l ll l l ll l l Cooccurrences W o r d N e t S i m il a r i t y agreement P ( x ) − − s ( i ) Fig. 2. (a) The average WordNet similarity for word pairs ( y -axis) as the number of pair co-occurrences ( x -axis)increases; (b) Agreement scores between pictures and cultural terms; (c) The silhouette value (the “goodness”of our taxonomy) at each creation step: from the second step with the Wikipedia taxonomy only ( wiki ), tothe third with the Flickr graph expansion ( flickr ), to the fourth which merged Flickr and Wikipedia ( aug ), tothe fifth which produced the validated and final taxonomy ( val ). Each box shows the four quartiles of thedistributions: the vertical lines indicate the top and bottom quartiles of values; the boxes are the mid-upperand lower quartiles, while the horizontal line in the middle shows the median value for the distribution.Outliers are shown as points on the graph. co-occurrences on Flickr pictures. We did so because past studies had shown that tags that oftenco-occur in the same photo are semantically related to each other [21]. We identified all the Flickrphotos that contain at least one of the 441 terms, paired them with all the co-occurring Flickr tags,and characterized each pair with the corresponding number of co-occurrences. By doing so, wefound 373,849 co-occurrences: our 441 terms co-occurred with 29,495 new unique tags. Step 4: WordNet Similarity Filtering.
Of course, most of those co-occurrences were semanticallyirrelevant. We discarded the irrelevant ones by removing all pairs of terms that occurred a numberof times less than a given threshold. To determine that threshold, we computed the similarity of termpairs as a function of different co-occurrence thresholds: from a number of co-occurrences of 100to one of 2500 (Figure 2a). The similarity of a pair of terms t , t was computed using WordNet [5]: sim path ( t , t ) = ∗ depth − len ( t , t ) (1)where len ( t , t ) is the shortest path distance between t and t in WordNet, and depth is themaximum distance between any two WordNet words. The higher the value, the more similar thetwo terms. As done in previous work [16], we computed the average path similarity ( sim path )between all pairs of terms retained for each threshold value. In Figure 2a, we see that the similarityconsiderably grows at first and then reaches a plateau at around a threshold value of 1800 co-occurrences. After it, the similarity still grows, but the corresponding standard deviations are toohigh. Therefore, conservatively, we kept all term pairs retained after applying a threshold of 2000co-occurrences ( n = Second ‘is-a’ filtering.
Not all the resulting 633 terms might have been related with ‘is-a’ relation-ships, as the construction of a taxonomy would require. To double check that, we explored each ofthese 633 terms and manually filtered out those that were not linked with ‘is-a’ relationships. Thissecond ‘is-a’ filtering resulted in 533 terms.
Step 5: Cleaning.
To make sure that all the 533 terms were relevant, we performed a final cleaningstep of potentially noisy terms. For each term, we drew a stratified random sample ( n =
50) of ig. 3. The wheel of our cultural taxonomy. The outer part shows examples of cultural terms (among the 263),the inner part shows the main 9 categories, and the middle part shows the 25 subcategories. pictures marked with that term. We then labeled each image as either being related to the term ornot. This made it possible to compute the average term’s “agreement” with its corresponding 50photos. We found that the majority of terms were in complete agreement with their photos anddid reflect cultural assets (Figure 2b). Conservatively, as a final step, we removed the terms thathad an agreement lower than 0.75. This resulted in 263 terms, which are the leaf nodes of our finalthree-level taxonomy (Figure 3). Validation.
The assumption behind the 5-step process was that each step resulted in a newset of terms that were better than the previous step’s set. To ascertain whether that assumptionwas true, we measured whether the set of terms in the same top-level category (we have nineof such categories) was cohesive (the terms in the same category were all related to each other)and distinctive (the terms in different categories were orthogonal to each other). To that end, wemeasured the clustering silhouette [22]. The silhouette s of a term i within cluster C (its top-levelcategory) determines how well i lies within C : s ( i ) = sim int ( i ) − sim ext ( i ) max { sim int ( i ) , sim ext ( i )} (2)where sim int ( i ) is the average path similarity (as per Formula 1) between term i ∈ C and anyother term j that is in the same cluster C ; conversely, sim ext ( i ) is measured by first computing,for each cluster C ′ (cid:44) C , the average similarity values between i ∈ C and all the terms in C ′ andthen selecting the highest average similarity. The values of s ( i ) range in [-1,1]: high values indicatecohesion, and low values indicate separation.We compared the distribution of silhouette values computed at each of the steps of taxonomycreation (Figure 2c): from step 2 ( wiki ) to step 5 ( val ). From the plot in Figure 2c, we see that the c u l t u r a l pho t o s LondonNew York 0.000.050.100.150.20 2010 2011 2012 2013 2014 % c u l t u r a l pho t o s LondonNew York
Fig. 4. Cultural content is consistently present over the five years under study: photos per annum (left) andfraction per annum (right).
Fig. 5. Cultural capital for neighborhoods in London (left) and New York (right). Neighborhoods are coloredin terms of the amount of cultural capital they possess. The top 25% of neighborhoods are depicted in lightblue, while the bottom are the darkest. median silhouette value indeed increases at each step: the median silhouette increases from 0.20 atstep 2 (where only Wikipedia terms are considered) to 0.40 in the final step.
Photos have been found to be good data sources for measuring people’s perceptions of public placesand for identifying distinctive features of the urban space (e.g., street art, temporary fairs) that aresurveyed neither by the census nor by open mapping tools [1, 13, 15, 17, 20, 21]. To trace culturalpatterns in our user-generated pictures, these pictures needed to be mapped onto geographical areasof interest. For London, we used its 33 boroughs; similarly, for New York, we used its 71 communitydistricts (60 of which qualified for our analysis due to lack of data for the others). We assignedeach picture to the corresponding census location l . To minimize the bias of our cultural profilingof cities towards amenities that are popular mostly among tourists, we filtered out non-locals byexcluding any Flickr user who had been active in each of the two cities for less than 30 days, ina way similar to previous work [9]. We then retained only pictures marked with at least one tagmatching one of the terms in our cultural taxonomy. This left us with 1 . M pictures. These picturescovered the period of five years with striking consistency (Figure 4). They also captured cultural itality across neighborhoods. To see why, consider that, as opposed to New York, in London,the official number of cultural venues by borough is made publicly available . We correlated thenumber of our cultural-related pictures with the number of cultural venues and found a Pearsoncorrelation coefficient as high as 0.70 ( p < . l , we computed the fraction of tags thatmatch any of the words in our taxonomy: f cult ( l ) = l l . (3)We then normalized those fractions using z -scores to obtain our estimate of cultural capital forlocation l : capital cult ( l ) = f cult ( l ) − µ ( f cult ) σ ( f cult ) , (4)where µ and σ are, respectively, the mean and standard deviation of the f cult distribution overall locations. Values of capital cult are displayed on the map in Figure 5. The values below zeroindicate locations with fewer cultural activities than those in the average location, while valuesabove zero indicate locations with greater cultural activities. Similarly, we computed an estimate ofthe economic capital of a location as: capital econ ( l ) = income ( l ) − µ ( income ) σ ( income ) , (5)where income ( l ) is the median income of resident taxpayers . The z -score represents the relativelyhigh or low culture (or economic capital) that is characteristic of a location in units of standarddeviations from the mean of the entire city. This allows us to draw an effective comparison notonly between areas but also between the two forms of capital. To estimate the cultural capitalunder a specific top-level taxonomical category c , we computed the relative presence of tags ofthat top-level category in the location of interest, normalized across locations: f cult ( l , c ) = c @ l l ; (6) capital cult ( l , c ) = f cult ( l , c ) − µ ( f cult ( c )) σ ( f cult ( c )) . (7)To mark locations with their most distinctive type of cultural asset, we defined the cultural speciali-sation of a location l as the category c with the highest capital: special cult ( l ) = arg max c ( capital cult ( l , c )) . (8)To capture how diverse a location is in terms of the variety of dimensions expressed throughcultural content, we computed its cultural diversity as the Shannon entropy over the category-specific cultural capital values within that location: diversity cult ( l ) = H cult ( l ) = − (cid:213) c f cult ( l , c ) × ln ( p cult ( l , c )) . (9)Shannon Entropy does not take into account the finite size of the sample: low sample sizes (withrespect to the number of bins) create biases towards higher entropy values. To overcome this Income (£) P London f cult ( l ) P London
Income ($) P New York f cult ( l ) P New York
Fig. 6. Distributions of variables encoding cultural and economic capital in London and New York. problem we apply the Miller-Madow’s correction to the entropy computation [18]. High diversityvalues indicate locations with cultural interests that span across the nine top-level categories, whilelow diversity values indicate locations with cultural interests specialized in a specific top-levelcategory.The goal of this study is to explore how the two forms of capital (cultural and economic)are linked to urban development. To meet that goal, we needed to collect metrics that captureurban development. The Index of Multiple Deprivation (IMD) for London boroughs and theSocial Vulnerability Index (SVI) for New York census tracts (denoted in the following as dev) hadbeen used as proxies for urban development before. Both are composite measures of deprivationacross several domains such as education, barriers to housing, crime, employment, and access toresources.We collected the IMD and SVI values for both years 2010 and 2014. The goal of this studyis to explore how the two forms of capital (cultural and economic) are linked to urban development.To capture cultural capital, we processed the pictures continuously from 2007 to 2015 as we havepreviously specified. Figure 6 shows the frequency distributions of cultural capital in both cities.To estimate economic capital, we gather data about London boroughsâĂŹ median income andmedian house prices released every year from 2007 and 2015, and New York’s tracts’ average incomeand median house prices released in 2014 (as only that year was publicly available for New York).Figure 6 shows the frequency distributions of income in London (in terms of British pounds) and inNew York (in terms of dollars). To capture urban development, we gather census data reportingthe London boroughs’ Index of Multiple Deprivation (IMD) released in both 2010 and 2014, andNew York tracts’ Social Vulnerability Index (SVI) released in both 2010 and 2014. Both reflecturban development as they are composite measures of deprivation across several domains such aseducation, barriers to housing, crime, employment, and access to resources. Both our taxonomyof urban culture and presence of its terms in each London borough and New York tract are madepublicly available under http://goodcitylife.org/cultural-analytics. Following the framework defined by Bourdieu in the context of social class and drawing an analogybetween social class and urban development, we ask if for neighborhoods, much like for people,cultural capital leads to positive development. As Bourdieu himself argued, prosperity cannotbe fully explained by economic capital alone. We therefore considered both cultural capital andeconomic capital of neighborhoods in 2010 and checked to what extent they predicted urbandevelopment (IMD in London and SVI in New York) five years later—at the beginning of 2015—through the following linear regression: dev = α + β · capital cult + β · capital econ (10) https://data.london.gov.uk/dataset/indices-deprivation-2010 SVI New York: http://data.beta.nyc/dataset/social-vulnerability-index or London, cultural and economic capital in 2010 are used to predict urban development in 2015.For New York, where less granular data is available, the average income in the period 2010-2014is used instead. We used ordinary least squares regression as a method to fit the data. Culturaland economic capital measured in 2010 are strong predictors of the development of an area in2015 in both cities (Table 1). In New York, the development score is relatively well explained byeconomic capital alone ( R = . ∆ dev ), we find that both types ofcapital have comparable roles in both cities. This suggests that improvement in neighborhoodsis a function of both economic capital and cultural capital. This is visually confirmed in Figure 7where each dot corresponds to a neighborhood, its position depends on the two values of capital forthat neighborhood, and its size reflects a positive change in development over the five years understudy. As opposed to what happens in London, in New York, improvements are more significantfor already economically prosperous neighborhoods. Despite this difference, cultural capital stillremains a powerful currency in both cities: positive changes in development scores are higheralong the cultural capital axis in both cities. When controlling for Flickr penetration (z-score ofnumber of tags in the neighborhood) the results change only slightly, with a relative increase in R between −
1% and + To see in what kind of activities cultural capital translates into, we explored which type of cultureis consumed in different neighborhoods. We measured the cultural characteristics of a location interms of the nine top-level dimensions of our taxonomy. We mapped the cultural specialization ofneighborhoods ( special cult , defined in Equation 8) in Figure 8 (left panels). In both cities, ‘Perfor-mance Arts’ appears in central areas, and ‘Architecture’ is predominant in either central areas andperipheral ones. As opposed to New York, London deviates from this typical pattern at times: EastLondon specializes in ‘Design’, and West London specializes in ‘Performance Arts’ and ‘Marketing’.To place these observations in context, we drew again the quadrant of economic capital vs. cultural capital (Figure 8, right panels), but, this time, the color of a node reflects the correspondinglocation’s specialization, and its size reflects the location’s cultural diversity ( diversity cult inEquation 9). In both cities, neighborhoods with high cultural capital do specialize in ‘PerformanceArts’ (in London, they do specialize in ‘Design’ and, to a lesser extent, in ‘Publishing and Media’ too).Also, neighborhoods with increasing urban development tend to be high not only in cultural capitalbut also in cultural diversity. The observation that higher urban development is associated with
Regression coefficients R α capital cult capital econ dev lon -0.51 3.4 ∗∗∗ -4.5 ∗∗∗ dev ny ∗∗∗ ∆ dev lon -0.18 2.26 ∗∗ -3.91 ∗∗∗ ∆ dev ny ∗∗ ∗∗∗ Table 1. Linear regression to model urban development dev (at the beginning of 2015) in terms of economiccapital and social capital. Regression coefficients, goodness of fit ( ∗∗ p<0.001; ∗∗∗ p<0), and the coefficient ofdetermination R are shown. ∆ dev is the difference between the development measured in 2015 and thatmeasured in 2010. l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l Barking and DagenhamBarnetBexleyBrentBromley CamdenCity of LondonCroydon Ealing Enfield Greenwich HackneyHammersmith and FulhamHaringeyHarrowHaveringHillingdon Hounslow IslingtonKensington and ChelseaKingston upon Thames LambethLewishamMerton NewhamRedbridgeRichmond upon Thames SouthwarkSutton Tower HamletsWaltham ForestWandsworth Westminster
Cultural Capital E c ono m i c C ap i t a l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l Greenwich VillageBattery Park Lower East SideMidtownClintonMurray Hill East HarlemLongwoodHunts PointTremontConcourseBelmont Bedford Park RiverdaleCastle HillCo−op CityPelham Parkway GreenpointBrooklyn Heights BedfordBushwickPark SlopeSunset Park Crown Heights SouthBay RidgeBorough Park Sheepshead Bay BrownsvilleEast FlatbushCanarsie Astoria Jackson Heights ElmhurstRidgewood FlushingBriarwood Richmond HillHoward BeachBayside Far Rockaway
Cultural Capital E c ono m i c C ap i t a l Fig. 7. Ranked cultural and economic capital in 2010. The size of each node represents the change indevelopment score (IMD for London and SVI in New York) between 2010-2015. The darker and larger anode is, the more improvement it has had over the five year period. cultural diversity is in line with what urbanists have claimed to be the main driver of neighborhoodprosperity: having diverse industries in geographical proximity [10, 14]. Indeed, by adding entropyas a regressor in our model in Equation 10, we improved our goodness of fit to R = .
41 for ∆ dev lon and R = .
71 for dev lon for London (8% and 9% improvement, respectively).
One of the main concerns that developing neighborhoods face is increasing house prices. If urbandevelopment is analogous to social mobility, then the house value of a neighbourhood can be com-pared to social class in Bourdieu’s terms. Therefore, similar to what we did for urban development,we used a linear regression to predict house prices from the cultural and economic capital with thefollowing regression: house _ price = α + β · capital cult + β · capital econ (11)For London, cultural and economic capital in 2010 are used to predict the housing prince in 2015. ForNew York, where less granular data is available, the cultural capital in 2010 and the economic capitalin the period 2010-2014 is used to predict the average house price in the period 2010-2014. Theresults suggest that, again, the ability to predict and explain house prices is not a purely economicmatter (Figure 9). Both forms of capital play a significant role in the model ( β = . , β = . R of 0 .
81 in London, and 0 .
56 in New York. Naturally, people choose to live in areas theycan afford and therefore economic capital still plays a fundamental role in explaining housingprices, however, an economic regressor alone achieves an R = .
53 in London, and R = .
48 inNew York, showing the importance of cultural capital. By then re-computing a regression for eachof the 9 top-level categories of cultural capital (Table 2), we explored whether a specific type ofcultural capital is associated with increasing house prices. In New York, ‘
Publishing ’ (e.g., contentlabeled as newspaper or books ) is most indicative of increasing housing prices, and a linear modelfor predicting house prices with it alone outperforms the composite cultural capital model ( R =0.59). rchitectureCraftsCultureDesignMarketingMediaPerformancePublishingTechnology l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l Barking and DagenhamBarnetBexleyBrentBromley CamdenCity of LondonCroydon Ealing Enfield Greenwich HackneyHammersmith and FulhamHaringeyHarrowHaveringHillingdon Hounslow IslingtonKensington and ChelseaKingston upon Thames LambethLewishamMerton NewhamRedbridgeRichmond upon Thames SouthwarkSutton Tower HamletsWaltham ForestWandsworth Westminster
Cultural Capital E c ono m i c C ap i t a l ArchitectureCraftsCultureDesignMarketingMediaPerformancePublishingTechnology l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l Greenwich Village
Battery Park
Lower East Side
Midtown
ClintonMurray Hill
East Harlem
Longwood
Hunts Point
Tremont
Concourse
Belmont
Bedford ParkRiverdaleCo−op City
Pelham Parkway
GreenpointBrooklyn Heights
Bedford
Bushwick
Sunset Park
Crown Heights
Bay Ridge
Borough Park
Brighton Beach
Sheepshead Bay
East Flatbush
Canarsie
Astoria
Sunnyside
Jackson Heights
Elmhurst
Ridgewood
Flushing
Briarwood
Richmond Hill
Bayside
Jamaica
Far Rockaway
Port Richmond
Cultural Capital E c ono m i c C ap i t a l Fig. 8. The cultural specialisation of neighbourhoods in London (top) and New York (bottom). On the left, themaps of the most representative cultural assets in different neighbourhoods; on the right, quadrants thatrelate cultural capital, economic capital, cultural specialisation (color of dots), and cultural diversity (size ofdots).
In London, ‘
Technology ’ is associated with increasing house prices ( R = . We have seen that cultural capital is associated with socio-economic development and increasinghouse prices. One might now wonder how cultural capital is generated. Since one simple way is ouseVal = + (cid:215) cult + (cid:215) econ , R = l l l ll l l l l l l l lll lll lll ll ll l l l l l ll ll l ll ll l l l l l l l l l l l l l l l l l l l ll ll l l l l l l l l l l l l l l l l l l l l l l lll lll lll l l l l l l l l l l ll ll ll ll l l l l ll ll ll ll ll ll l l l l l ll ll l l l l l l l l l ll l l l l l l l l l l Lambeth WestminsterBexley IslingtonBarking−2.00−1.75−1.50−1.25−1.00−0.75−0.50−0.250.000.250.500.751.001.251.501.75 − . − . − . − . − . − . − . − . . . . . . . . . predicted a c t ua l houseVal = + (cid:215) cult + (cid:215) econ , R = l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l ll ll Concourse Riverdale Forest HillsQueens VillageRidgewoodBaysideBorough Park Brooklyn HeightsMurray HillMidtownEast HarlemLower East Side−1.75−1.50−1.25−1.00−0.75−0.50−0.250.000.250.500.751.001.251.501.752.002.252.502.75 − . − . − . − . − . . . . . . . . . . predicted a c t ua l Fig. 9. Linear regression results for housing price z -scores across neighbourhoods over the period 2010-2015.The regression line is shown in red and the shaded area around it represents the limits of the 95% confidenceinterval. London NYArchitecture 0.71 0.52Crafts 0.63 0.55Culture 0.64 0.58Design 0.62 0.57Marketing 0.62 0.5Media 0.63 0.51Performance 0.64 0.52Publishing 0.64
Technology
Table 2. R coefficients for different types of cultural activities in predicting housing prices. through cultural events, we set out to detect such events in our our digital data. To detect peaks inthe fluctuation of the cultural capital that might correspond to key cultural events, we measuredthe cultural capital of a neighborhood on a running monthly basis and compared it to the expectedvalue in that neighborhood. More specifically, we computed the z -score of the fraction of culturalcontent at month t ∈ [ , T ] using the average and standard deviation of the fraction measured inall months (0 to T ) at location l : capital tcult ( l ) = f tcult ( l ) − µ ( f [ − T ] cult ( l )) σ ( f [ − T ] cult ( l )) . (12)Figure 10 (left) shows the variation of the cultural capital over time for the five neighborhoodsin London and New York that had the highest variation of urban development ( ∆ dev ) between2010 and 2015. Peaks and falls are easy to see, if contrasted to the horizontal line (which is theneighborhood’s (typical) mean cultural capital). For each of the top outliers in the neighborhoods in ast Harlem Greenpoint Greenwich Village Midtown SunnysideGreenwich Hackney Newham Tower Hamlets Waltham Forest −2024−2024 z − sc o r e ll ll ll ll ll l l ll ll l ll ll East HarlemGreenpointGreenwich VillageMidtownSunnysideGreenwichHackneyNewhamTower HamletsWaltham Forest−2 0 2 4 monthly cultural capital
Fig. 10. Cultural capital for the five neighborhoods in London and New York that most improved over the fiveyear period (2010-2015). Left: inter-borough z -scores of cultural capital on a monthly basis shows divergencesfrom the mean (purple line). Right: Distribution of cultural capital monthly values across neighborhoods(outliers are events with considerably higher cultural capital). l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l ll l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l ll l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l l l l l l l l ll l l l l l l l Barking Barnet Bexley Brent Bromley Camden City of London Croydon Ealing Enfield GreenwichHackney Hammersmith Haringey Harrow Havering Hillingdon Hounslow Islington Kennsington Kingston LambethLewisham Merton Newham Redbridge Richmond Southwark Sutton Tower Hamlets Waltham Forest Wandsworth Westminster z − sc o r e ll EconomicCultural
Fig. 11. Evolution of cultural and economic capital for different London boroughs in the period 2007-2014.Because the values are z-score normalized they are comparable: a borough with an economic capital higherthan the social capital (or viceversa) indicates means that, relative to all other neighborhoods, that boroughis better in terms of its economic rather than cultural status. Curves crossing indicate that one type of capitalbecomes more prominent than the other, relatively to all other neighborhoods.
Figure 10, we identified the exact event that took place. In Table 3, we see that a variety of culturalevents were indeed at the heart of changing and enhancing the reputation of specific places in bothcities.Based on a wider temporal analysis in London (Figure 11), one can see that cultural capitaltranslates into economic capital in a few years. Areas subject to cultural revitalization eventuallygentrify [6, 25].
Culture pays. That is not always obvious for policy makers. “When budgets have come underpressure, there has been a tendency for arts and culture to be viewed as ‘nice to have’, rather than anecessity” [12]. Culture surely comes with intrinsic benefits: it opens our minds to new emotional ate event borouдh photos / users cult cateдory imaдe Table 3. Summary of top events in the five neighborhoods of London and New York that most improved from2010 to 2015. For each event, the table reports its date, name, location, number of users involved in it, thechange it caused in cultural capital in terms of deviation from the mean ( cult ), the most frequently occurringcultural category, and a representative image. experiences, and enriches our lives. But, as we have shown, it also comes with extrinsic benefits: it isa catalyst for positive change and growth in neighborhoods. We have found that the neighborhoodsin London and New York that experience the greatest growth are those with high cultural capital.The production of these findings relies on a new way of quantifying cultural capital that is based n the definition of the first taxonomy of culture (which is far more comprehensive than officialclassifications of cultural activities) and on the mining of digital data such as picture tags (whichhas made it possible to perform cultural studies at an unprecedented scale, contributing to theemergence of a new research field called ‘ cultural analytics ’ [15]). Despite being geographicallybiased, picture data has been valuable not only for observing neighborhood growth and identifying“up-and-coming” areas but also for predicting house prices.Culture pays, but only up to a point. As Bourdieu argued, cultural inequality widens and le-gitimizes economic inequality [11]. As such, culture—which powers the growth of cities—alsocauses their distressing challenges: gentrification, unaffordability, and inequality [8]. A sustainableapproach to cultural investments might pay dividends but requires sensitivity to the needs of localcommunities. REFERENCES [1] Aiello, L. M., Schifanella, R., Quercia, D., and Aletta, F. (2016). Chatty maps: constructing sound maps of urban areasfrom social media data.
The Royal Society Open Science
3, 150690[2] Bakhshi, H., Freeman, A., and Higgs, P. L. (2012). A dynamic mapping of the UK’s creative industries.
Nesta [3] Bourdieu, P. (1984).
Distinction: A social critique of the judgement of taste (Harvard University Press)[4] Evans, G. (2009). Creative cities, creative spaces and urban policy.
Urban Studies
46, 1003–1040[5] Fellbaum, C. (1998).
WordNet (Wiley)[6] Florida, R. (2003). Cities and the creative class.
City & Community
2, 3–19[7] Florida, R. (2014).
The Rise of the Creative Class (Basic Books)[8] Florida, R. (2017).
The New Urban Crisis: How Our Cities Are Increasing Inequality, Deepening Segregation, and Failingthe Middle ClassâĂŤand What We Can Do About It (Basic Books)[9] Girardin, F., Calabrese, F., Dal Fiore, F., Ratti, C., and Blat, J. (2008). Digital footprinting: Uncovering tourists withuser-generated content.
IEEE Pervasive computing
7, 36–43[10] Glaeser, E. (2011).
Triumph of the City: How Urban Spaces Make Us Human (Pan Macmillan)[11] Grenfell, M. (2014).
Pierre Bourdieu: Key Concepts (Taylor & Francis)[12] Henley, D. (2016).
The Arts Dividend: Why Investment in Culture Pays (Elliott & Thompson)[13] Hristova, D., Williams, M. J., Musolesi, M., Panzarasa, P., and Mascolo, C. (2016). Measuring urban social diversityusing interconnected geo-social networks. In
Proceedings of the th ACM Conference on World Wide Web (WWW) .21–30[14] Jacobs, J. (1969).
The Economy of Cties (Random House)[15] Manovich, L. (2009). Cultural analytics: visualising cultural patterns in the era of more media.
Domus [16] Meng, L., Huang, R., and Gu, J. (2013). A review of semantic similarity measures in WordNet.
International Journal ofHybrid Information Technology
6, 1–12[17] Noulas, A., Scellato, S., Mascolo, C., and Pontil, M. (2011). Exploiting Semantic Annotations for Clustering GeographicAreas and Users in Location-based Social Networks. In
Proceedings of the th AAAI Conference on Web and SocialMedia (ICWSM) [18] Paninski, L. (2003). Estimation of entropy and mutual information.
Neural computation
15, 1191–1253[19] Ponzetto, S. P. and Strube, M. (2007). Deriving a Large Scale Taxonomy from Wikipedia. In
Proceedings of the nd Conference on Artificial Intelligence (AAAI . vol. 7, 1440–1445[20] Quercia, D., Schifanella, R., and Aiello, L. M. (2014). The shortest path to happiness: Recommending beautiful, quiet,and happy routes in the city. In
Proceedings of the th ACM Conference on Hypertext and Social Media . 116–125[21] Quercia, D., Schifanella, R., Aiello, L. M., and McLean, K. (2015). Smelly Maps: The Digital Life of Urban Smellscapes.In
Proceedings of the th AAAI Conference on Web and Social Media (ICWSM) [22] Rousseeuw, P. J. (1987). Silhouettes: a graphical aid to the interpretation and validation of cluster analysis.
Journal ofComputational and Applied Mathematics
20, 53–65[23] Törnqvist, G. (2011).
The geography of creativity (Edward Elgar)[24] Yencken, D. (1988). The creative city.
Meanjin
47, 597[25] Zukin, S. (1987). Gentrification: culture and capital in the urban core.
Annual review of sociology
13, 129–14713, 129–147