Neighbourhood change among creative–cultural districts in mid-sized US metropolitan areas, 2000–10

Abstract

Cities across the United States are experiencing a surge in urban residential development, particularly within the downtown core and other centrally located, historic, mixed-use urban neighbourhoods. Often situated in former warehouse districts or other ‘marginal’ areas, many of these neighbourhoods have been revalorized and reinvented as hubs of creative–cultural production and consumption. Through an examination of 102 US neighbourhoods in 70 mid-sized metropolitan areas, this article explores patterns in the physical form, geography, and recent demographic and socio-economic evolution of the ‘creative–cultural district’ (CCD). The results of this exploratory analysis suggest that over the previous decade (2000–10), the majority of CCDs attracted high-skill, high-wage creative–knowledge workers at a rate faster than their respective metropolitan areas. This ‘creative gentrification’ was also evident in rising residential populations, household income, education and rent. Rent increased faster than income in over half the surveyed neighbourhoods, suggesting that a widening deficit in affordable housing has accompanied growing consumer demand for vibrant urban neighbourhoods and associated arts/cultural/entertainment amenities.

Keywords:

• cultural district
• creative class
• creative city
• neighbourhood change
• gentrification
• creative gentrification

Introduction

While dynamic, creative–cultural or ‘neo-bohemian’ (Lloyd, 2002) urban neighbourhoods have thrived for decades in the United States’ premier world cities, such as New York, Los Angeles, San Francisco and Chicago, the scope of their recent success and popularity in cities across the country, as well as their propensity to attract a sizable residential population, is unprecedented in recent times. Although the issues of displacement, inequity and homogenization associated with gentrification continue to present considerable challenges, the revitalization of urban neighbourhoods may represent the stirrings of a significant and broadly supported turnaround in the economic fate of many US cities. This may be particularly true of mid-sized US metropolitan areas (roughly those with populations between 300,000 and 3.5 million), many of which still struggle to maintain healthy, liveable and relevant urban centres. This article explores the creative–cultural district (CCD) as it exists in the United States, building upon and extending previous examinations of creative–cultural neighbourhoods (Evans, 2009; Lloyd, 2002; McCarthy, 2005; Montgomery, 2003) and investigating the physical form, geography, and recent demographic and socio-economic shifts among a large (N = 102) and diverse sample of urban neighbourhoods.

Table

	CCD	Historic district	BID	Walk score	Special district^a	% change in the population, 2000–10	% 20–34, 2000	% 20–34, 2010	% 62+, 2000	% 62+, 2010	% BA 25+, 2000	% BA 25+, 2006–10	% Creative class, 2000	% Creative class, 2006–10	% Change in income, 1999–08	% Change in gross rent, 2000–08	Cluster
Akron, OH	South Main Street	×		85		155	41	19	17	9	10	17	25	34	–14	89	2
Albany, NY	Downtown Albany HD	×	×	94		–7	35	20	26	24	16	17	21	30	21	38	2
	Lark Street	×	×	91		–1	68	42	9	11	58	68	53	63	6	45	1
Albuquerque, NM	Central Avenue/Downtown	×		80	A/CD	9	69	39	6	6	16	25	53	58	–8	17	1
	Nob Hill	×		77		1	60	32	7	9	52	49	44	49	–9	22	2
Asheville, NC	Downtown	×		100		3	54	26	21	24	19	33	26	35	58	84	2
Augusta, GA	Broad Street HD	×		86		–5	65	17	18	23	64	72	46	65	9	94	1
Austin, TX	Warehouse District	×		100		131	57	46	14	8	43	64	50	69	28	110	3
Baltimore, MA	Mt. Vernon	×	×	97	CD	12	62	44	15	14	45	59	50	26	16	68	1
	Little Italy			91		102	53	39	19	12	32	36	46	37	25	122	1
	Fell’s Point	×		90	EZ	36	58	47	14	12	57	69	65	69	–4	94	1
Birmingham, AL	Loft District	×	×	85		10	49	29	12	11	5	19	30	51	61	132	2
	Five Points South	×		89	EZ	14	61	28	8	6	26	35	37	38	–18	40	2
Charleston, SC	King Street	×		95		15	44	10	5	5	34	48	27	30	68	67	2
Charleston, WV	Capitol Street	×		83		–12	34	22	26	23	18	11	28	22	15	79	2
Chattanooga, TN-GA	North Shore	×		74		9	43	30	19	18	40	56	43	54	28	75	1
Cincinnati, OH-KY	Over the Rhine	×	×	94		–9	43	27	7	7	8	24	17	46	8	81	2
	Covington	×	×	89	CD	6	48	32	21	19	23	36	37	42	241	37	3
Cleveland, OH	The Flats/Warehouse District	×	×	86	ED/EZ	78	68	42	4	3	27	41	55	62	17	3	1
Colorado Springs, CO	North Downtown		×	89		13	56	15	7	6	55	40	46	37	0	22	2
	Old Colorado City	×		94		–9	41	24	14	17	22	27	33	38	–11	–1	2
Columbia, SC	Congaree Vista	×		85		76	50	39	9	5	37	66	44	32	22	45	1
	Five Points			72		–6	47	22	13	16	45	49	45	47	–7	43	2
Columbus, OH	Short North	×	×	80	AD	27	66	56	7	4	45	73	44	60	38	76	1
Dayton, OH	Oregon HD	×	×	84	AD	–1	33	16	16	19	8	10	26	32	–16	47	2
Daytona Beach, FL	Beach Street/Downtown	×		83		–20	38	10	18	17	15	24	20	18	–5	98	2
Denver, CO	LoDo	×	×	91	EZ	96	38	40	28	22	50	69	59	68	71	233	3
Des Moines, IA	Court Avenue District			89	CD/ED	9	41	32	25	13	17	44	28	54	23	59	3
Eugene, OR	Kesey Square/Broadway			94		5	64	35	14	15	39	41	38	30	–11	32	2
Fort Myers, FL	River District	×		86		–12	44	24	27	31	16	21	32	28	9	89	2
Grand Rapids, MI	Heartside	×	×	87		99	26	9	38	18	16	20	30	35	–20	78	2
Greensboro, NC	South Elm Street	×	×	82		69	39	30	31	17	43	40	32	57	–25	2	2
Greenville, SC	Main Street			91		77	39	36	36	20	38	64	49	65	110	150	3
Indianapolis, IN	Massachusetts Avenue	×		95	CD	57	53	37	14	10	39	47	53	60	2	32	1
	Broad Ripple			80	EZ/CD	15	52	38	15	13	66	69	54	59	–26	17	1
Jacksonville, FL	Five Points/Riverside	×		95		–5	41	29	25	25	21	37	30	44	11	62	2
	San Marco Square			72		1	40	20	13	15	45	45	48	49	–25	67	1
Kansas City, MO-KS	Crossroads			89	AD	248	67	54	2	3	7	44	20	63	–	–4	3
	Westport			80	ED	1	48	28	14	14	44	40	44	42	14	32	2
Knoxville, TN	Old City/Market Square	×	×	91		23	61	37	16	10	22	44	43	62	82	128	3
Lexington, KY	South Hill	×		78		79	59	16	6	3	25	32	30	33	–23	70	2
Lincoln, NE	Haymarket	×	×	84		6	59	19	12	12	40	43	34	51	40	113	2
Little Rock, AR	River Market		×	85		38	53	33	15	11	25	20	33	37	28	25	2
Louisville, KY-IN	East Market District/NuLu	×		91		–8	40	25	12	10	15	17	28	30	28	79	2
	West Main District/Downtown	×		72		67	48	34	20	10	16	22	40	41	38	61	2
Memphis, TN	South Main/Beale Street	×	×	89	AD	9	46	42	23	10	42	57	51	51	20	33	1
Milwaukee, WI	Historic Third Ward	×	×	89	AD/EZ	378	58	48	3	5	61	61	63	54	11	42	1
	Old World 3rd Street	×	×	94		34	75	41	2	3	22	34	52	48	17	22	1
	Water Street	×	×	95	ED/EZ	13	60	34	14	13	57	68	60	58	–2	6	1
Minneapolis, MN	Warehouse District	×	×	91	EZ	183	62	46	3	5	40	77	51	73	57	95	3
	Uptown			98		3	54	46	5	6	51	60	43	57	1	26	1
Mobile, AL	LoDa/Lower Dauphin	×	×	86		5	32	21	35	26	22	31	49	55	28	–4	2
Naples, FL	5th Avenue South	×	×	78		–17	21	7	41	52	46	49	48	42	7	44	1
Nashville, TN	The District	×	×	97		122	72	42	4	6	24	39	40	66	48	33	1
New Haven, CT	9th Square	×	×	98		26	67	50	4	6	57	65	49	74	47	62	1
New Orleans, LA	French Quarter	×	×	94	CD/EZ	–5	45	26	17	24	51	51	39	53	33	51	1
	Warehouse District	×	×	85	AD	27	48	33	20	15	40	60	22	73	5	46	3
	Lower Garden District	×		77	CD	31	56	45	8	7	45	64	50	56	30	148	3
Oklahoma City, OK	Bricktown		×	66	ED	–60	30	20	10	19	15	23	12	11	15	118	1
Omaha, NE	Old Market	×	×	91		32	66	45	6	5	28	41	42	64	10	64	2
Pensacola, FL	Palafox Place/Downtown	×	×	89		–17	35	19	25	20	17	24	27	34	–14	81	1
Pittsburgh, PA	South Side Flats	×		88		15	48	35	25	14	25	46	42	52	33	36	2
	Strip District			83	EZ	132	34	45	14	9	34	64	32	54	14	153	1
	Oakland	×	×	81		23	73	6	6	4	51	45	38	31	–21	75	3
	Cultural District		×	100	CD	–31	61	20	12	14	23	31	54	61	58	40	1
Portland, ME	Old Port/Arts District		×	97	AD	–16	60	32	11	14	33	32	34	39	–32	33	2
Portland, OR	Old Town Chinatown	×	×	94	EZ	119	60	36	6	12	29	54	45	56	19	69	2
	Pearl District	×		95		137	62	40	6	11	35	56	48	60	49	143	1
	Northwest District/Nob Hill	×		92		9	66	47	12	13	49	64	49	57	–3	38	3
Providence, RI	Federal Hill	×		86		–4	51	32	11	8	19	45	23	44	35	69	1
	Downtown	×	×	93		71	38	17	19	11	36	37	30	46	–19	68	2
	College Hill/Thayer Street	×		90		6	69	18	6	5	77	77	68	66	23	60	3
Raleigh, NC	Warehouse District/Downtown	×	×	88		53	57	35	16	11	28	39	42	50	–29	48	1
	Glenwood South	×		75		23	50	38	22	18	55	60	58	65	19	–4	1
Richmond, VA	Shockoe Bottom	×		77		70	52	46	15	5	45	49	46	54	47	66	1
Roanoke, VA	City Market	×		92		38	75	42	3	4	–	31	10	69	169	153	1
Sacramento, CA	Old Sacramento	×	×	91		6	65	36	18	16	8	9	64	49	–7	66	3
San Antonio, TX	River Walk	×		98		2	51	33	19	20	16	25	30	42	21	49	2
San Diego, CA	Gaslamp Quarter	×	×	95	EZ	52	52	37	18	17	32	48	46	58	26	79	2
	Little Italy	×	×	92		208	54	54	8	7	29	74	40	61	127	171	1
San Jose, CA	Willow Glen		×	88		7	42	21	14	15	26	48	59	56	125	174	3
	Downtown/SoFa	×	×	98	A/ED	32	55	34	12	12	33	40	36	42	1	48	1
Santa Barbara, CA	State Street/Downtown	×	×	98	AD	1	44	26	20	20	29	29	31	30	–22	39	3
Savannah, GA	River Street	×		94		–14	55	22	7	10	53	53	41	53	57	45	2
Seattle, WA	Pike Place Market	×	×	100		17	46	31	16	18	35	48	60	69	11	121	1
	Pioneer Square	×	×	86	EZ	54	46	30	25	20	21	27	32	32	–6	72	1
	Belltown		×	98		50	56	43	14	11	50	63	50	62	29	60	2
	Pike/Pine		×	98		7	70	48	9	9	48	61	47	59	16	37	1
Springfield, MA	Metro Center/Club Quarter	×	×	82		6	49	27	8	7	12	9	26	25	–5	54	1
Springfield, MO	Walnut Street/Downtown	×	×	89		4	63	27	10	7	21	24	22	32	44	36	2
St. Louis	Downtown West	×		88		36	36	45	37	7	15	46	30	47	94	32	2
	Soulard	×		73		10	57	42	10	9	42	59	50	52	1	48	1
	Laclede’s Landing/Arch	×	×	85	EZ	226	49	43	13	6	40	54	40	56	2	37	3
Syracuse, NY	Armory Square	×	×	83	CD	–12	62	42	9	7	29	43	49	65	2	72	1
Tampa, FL	Ybor City	×		79	EZ	19	30	33	24	16	5	33	15	43	115	114	1
Toledo, OH	Warehouse District	×	×	77		–11	34	22	8	7	2	4	8	7	–38	14	3
Tucson, AZ	Iron Horse/Fourth Avenue	×	×	90		7	64	28	6	6	48	54	42	48	–9	44	2
Tulsa, OK	Blue Dome		×	86		–1	34	19	19	19	69	69	63	53	13	70	2
Norfolk, VA	Granby Street/Downtown	×	×	88		36	63	37	12	15	35	43	65	75	12	59	1
Wichita, KS	Old Town	×		71		3	65	34	5	6	17	19	32	31	–18	25	2
Winston-Salem, NC	DADA	×		95	AD	37	68	44	8	10	15	25	33	58	30	40	2
Worcester, MA	Green Island			89		7	41	25	14	9	3	14	11	9	18	35	2

Note:

^a Special district abbreviations: arts district (AD), cultural district (CD), entertainment district (ED), and entertainment zone (EZ).

CCDs may be described as centrally located, and often historic, urban neighbourhoods that serve as nodes of creative–cultural production and consumption (e.g., recording studios and music venues, art studios and galleries, live theatre, dining and retail, festivals and sporting events). Frequently located near or adjacent to the central business district (CBD), most CCDs in the United States are characterized by 19th- and early 20th-century architecture and urban form, with ‘traditional’ urban features such as multi-unit housing, a mix of residential and commercial land uses, pedestrian-oriented streetscapes, and access to public transit. Many CCDs are evolving rapidly, gaining residents through the conversion and reuse of old buildings and/or new construction. At the forefront of the ‘re-urbanization’ trend are young singles (particularly members of Generation Y or the ‘Millennials’, born 1977–94) and childless married couples, who are exercising their consumer preference for more vibrant, convenient, and ‘authentic’ urban neighbourhoods and lifestyles (Birch, 2005; Glaeser & Gottlieb, 2006; Urban Land Institute, 2013; Zukin, 2008). The growing appeal of these neighbourhoods among well-educated ‘creative’ or ‘knowledge’ workers have, over the last two decades, precipitated significant demographic and socio-economic shifts, in a process Peck (2005, p. 751) has described as ‘creative gentrification’.

Much of the literature on CCDs to date has focused on a single city, or a limited set of cities, to examine critically the nature and process of neighbourhood change at specific times and in specific places (Catungal, Leslie, & Hii, 2009; Tremblay & Battaglia, 2012; Zimmerman, 2008). This paper presents a broad overview of the CCD in the United States with the primary goal of elucidating more general patterns in their recent physical and socio-economic transformations, and to speculate on the potential implications of these changes. The specific objectives of this exploratory analysis were threefold: (1) to compile a list of central, urban neighbourhoods within the United States that exhibit concentrations of creative (e.g., art, knowledge) and cultural production and consumption, and, in most cases, those that have also played a key role in urban regeneration and place-making efforts; (2) to identify common features related to the physical form and activity patterns of these neighbourhoods; and (3) to assess the demographic and socio-economic profiles of CCDs over the previous decade to quantify the extent, direction and rate of recent neighbourhood change. This information will provide a quantitative, neighbourhood-level perspective on the nature and scale of urban revitalization and gentrification in mid-sized US metropolitan areas, and help identify potential implications of these changes amid the latest economic recession and recovery.

The remainder of the article is organized as follows. The first section provides an overview of the literature concerning the preservation and redevelopment of CCDs, and how their recent resurgence parallels growing demand for urban amenities, particularly among the creative–knowledge class. This is followed by a detailed look at the physical and social features common to CCDs based on a nationwide survey. Finally, a preliminary analysis is preformed to characterize the nature and degree of recent gentrification and neighbourhood change among 102 CCDs located in 70 mid-sized US metropolitan areas.

Background and context

The 1960s and 1970s were a particularly challenging time for many US cities. A severe decline in prosperity wrought by a shrinking manufacturing sector and an explosion in automotive-based suburban sprawl resulted in rising poverty and crime rates in the inner city (Glaeser & Gottlieb, 2006; Short, Benton, Luce, & Walton, 1993). While the impoverished became concentrated in dangerous housing projects and tax-starved inner-city neighbourhoods, large-scale urban renewal projects involving the construction of freeways, civic buildings, arenas and other public projects, gutted much of what remained of the physical and social cohesion of traditional urban neighbourhoods (Bauman, Biles, & Szylvian, 2000). At the same time, however, historic preservation and the value of the ‘urban village’, so famously portrayed and defended by Jane Jacobs in The Death and Life of Great American Cities (1961), began to impart significant influence on urban planning and policy (Klemek, 2008; Petrus, 2003; Zukin, 2010). Many CCDs owe their existence to such efforts by historical preservationists and other urban activists to preserve historic, human-scaled urban environments in central city locations.

In the final two decades of the 20th century, the urban core of many large US cities began to experience a rebound in residential population as violent crime dropped precipitously and the demand for urban amenities grew in step with rising wages and education (Birch, 2005; Ehrenhalt, 2012; Glaeser & Gottlieb, 2006; Wilson, Plane, Mackun, & Fischetti, 2012). Today, vibrant urban neighbourhoods with a unique historic heritage and distinct personality are commonly regarded as key cultural assets, capable of attracting human and financial capital (Dinnie, 2011; Gotham, 2007; Kavaratzis & Ashworth, 2007; Ponzini & Rossi, 2010). While cities and metropolitan areas still compete to lure corporate firms to spur economic growth, many have also begun appealing directly to the ‘talent’, most notably a cohort of educated professionals Richard Florida (2002, 2012) terms the ‘creative class’. Representing a diverse range of knowledge- and culture-intensive occupations from artists to engineers, members of the creative class are defined by their propensity to ‘create new forms’, whether they be oil paintings, fashion designs, architectural plans or academic manuscripts. Based on interviews and focus groups, Florida concluded that members of the creative class are particularly sensitive to ‘quality of place’ (Florida, 2012, pp. 280–303). Under the creative city paradigm, cities and neighbourhoods that exhibit diversity and tolerance, promote social interaction while also facilitating anonymity, and more broadly generate a creative milieu, are particularly prized by ‘creatives’ (Florida, 2012; Landry, 2008). Despite limited evidence that creative individuals are more likely to relocate based primarily on their perception of quality of place (as opposed to economic opportunities, family ties, etc.; Borén & Young, 2013), many cities have taken Florida’s concept to heart, incorporating the themes and language of the creative class into their branding and promotional strategies (Ward, 2010; Zimmerman, 2008).

Many who would likely identify as members of the creative class, including artists and well-educated young professionals, have exhibited a preference for urban neighbourhoods and lifestyles (Lawton, Murphy, & Redmond, 2013; Zenker, 2009). Artists have often been among the first to repurpose former industrial buildings, using them as studio space and live–work units (Markusen, 2006). In doing so, they play a pivotal role in transforming the image of entire city districts from marginal, decrepit, and dangerous to lively, hip and trendy (Hannigan, 2007; Lloyd, 2002). Embracing their role in urban economic development, clusters of art studios and galleries are frequently recognized by local governments as official arts or cultural districts (Chapple, Jackson, & Martin, 2010; Markusen & Schrock, 2006). Where high concentrations of workers are engaged in creative, cultural or knowledge-intensive occupations (e.g. information technology, architecture and engineering; physical and social science; arts and design; medicine; law; entertainment and media), amenities such as independent restaurants and retail stores, street cafés and quirky coffee shops often follow, establishing new spaces of consumption as well as common ground for networking and other social interactions – the supposed catalysts of innovation and economic growth (Chapple et al., 2010; Comunian, 2011; Lloyd, 2002; McGranahan & Wojan, 2007).

Some critical of Florida’s creative class thesis have argued that it is, at best, a new perspective on an old idea, and, at worst, a prescription for exacerbating intra-urban inequality. Peck (2005) has argued that, far from offering a new paradigm in urban economics, the creative class and creative city theses align seamlessly with the orthodox entrepreneurial policies and neoliberal strategies that have been embraced by pro-growth urban elites since the 1970s. The benefits of such policies, it is argued, are likely to accrue mainly to the affluent, including the ‘esteemed dignitaries of the creative class’ (Maliszewski, 2003, p. 77), while low-income groups and those unable to participate fully in the new creative–knowledge economy are further marginalized (McCann, 2007). Despite the potential for deepening social fragmentation and inequality, policy-makers have been unable to ignore the integral role creative, cultural and knowledge-based industries – broadly defined – now play in the modern, globalized economy (Atkinson & Easthope, 2009; Scott, 2006). Nurturing vibrant urban streetscapes, a tolerant and social people climate, and clusters of creative and cultural amenities are viewed as a means of attracting skilled labour and promoting economic growth (Boschma & Fritsch, 2009), though several authors have commented that CCDs are likely a product, as well as a driver, of economic success (Malanga, 2004; Markusen, 2006; Storper & Scott, 2009).

It is against this backdrop of on-going controversy regarding the potential merits and negative externalities of ‘creative gentrification’ and creative–cultural urban ‘reinvention’ that this study explores the nature and scale of recent neighbourhood change across a large number of US cities and their CCDs. Although much has been written of the rise of cultural districts and creative/culture-led gentrification, the research in this area has heretofore been primarily qualitative or limited to specific case studies (e.g., Catungal et al., 2009; Gotham, 2005; Ley, 2003; Stern & Seifert, 2010; Tremblay & Battaglia, 2012; Zimmerman, 2008; Zukin & Braslow, 2011). Though other studies have examined neighbourhood change across multiple cities (Farrell & Lee, 2011; Wei & Knox, 2013), the broad scope of their analyses, generally involving a wide range of neighbourhood typologies, has yet precluded a detailed discussion of CCDs specifically. In this study, a detailed empirical analysis is undertaken to gauge how CCDs are changing demographically and economically in relation to their respective metropolitan areas, and to investigate the potential role of special district designations, amenities and other neighbourhood features in catalysing these changes.

Identifying CCDs and their common features

A nationwide survey was first conducted to identify US neighbourhoods that may serve as exemplars of the CCD, and to elucidate further common characteristics related to the form and function of CCDs. This information is used in the section that follows to explore whether these features may have played a role in shaping the nature and pace of recent neighbourhood change. CCDs were identified on the basis of (1) having an official designation as either an arts, cultural or entertainment district, and/or (2) sharing several of the place characteristics outlined by Montgomery (2003), which included aspects of the neighbourhood’s activity patterns (i.e., the presence of cultural events and amenities, café culture and nightlife), built form (i.e., fine-grain morphology, mix of uses, variety in the age and types of buildings), and meaning (i.e., historical heritage, sense of place or identity). Information about each neighbourhood was gathered from a variety of sources including the primary literature, numerous books and periodicals, news articles, websites and by consulting, on an informal basis, experts knowledgeable of local conditions. The term ‘neighbourhood’ is used here to refer to places that often have fuzzy, if not contested, edges (Spielman & Logan, 2013), though in general have both a unique identity and an identifiable commercial core.

In all, 102 neighbourhoods were identified from an analysis of 70 mid-sized US metropolitan areas with populations between 300,000 and 3.5 million (see Appendix A). Given the particularly critical role arts, cultural and entertainment districts now play in the urban revitalization and rebranding strategies of cities perched among the middle-to-lower rungs of the world city hierarchy (Globalization and World Cities (GaWC), 2012; Gotham, 2007; Hagerman, 2007; Long, 2009; Ponzini & Rossi, 2010; Zimmerman, 2008), this paper focuses specifically on mid-sized US metropolitan areas. Note that the term ‘mid-sized’, as defined in this study, represents an attempt to balance the variability introduced by cities of different sizes with the desire to conduct a broad and inclusive neighbourhood survey. A different sample of cities with an alternative range of city sizes may yield different results. Although an effort was made to identify at least one CCD in every mid-sized US metropolitan area, the current list remains far from exhaustive, and omissions may be particularly acute among the larger metropolitan areas.

Special district designations

As a common element among urban redevelopment strategies, cities have acted to protect, revalorize, and promote unique historical and cultural assets through the sanctioning of historic, cultural, and entertainment districts (Fitch, 1990; Frost-Kumpf & Helmke, 1998; Mizzau & Montanari, 2008; Zukin, 2010). CCDs are often associated with one or more of these specialized districts (see Appendix A). Whether formal or informal, specialized districts often contribute significantly to the form, function and image of CCDs as hubs of creative–cultural activity. Cultural districts are defined as ‘well-recognized, labeled, mixed-use area[s] of a city in which a high concentration of cultural facilities serves as the anchor of attraction’ (Frost-Kumpf & Helmke, 1998, p. 7). Cultural facilities include concert halls, cinemas, theatres, libraries, museums, galleries and music venues. Closely allied terms include cultural clusters, which are ‘concentrations of cultural resources’ that contain ‘nonprofit arts organizations, commercial cultural firms, resident artists, and cultural participants’ (Stern & Seifert, 2010, p. 262) and the more comprehensive cultural district, which Montgomery (2003) characterizes as city districts with ‘cultural activity [… that] should include cultural production (making objects, goods, products, and providing services) as well as cultural consumption (people going to shows, visiting venues and galleries)’ (p. 296). Clusters of cultural assets that include a preponderance of art-related galleries, studios, institutions and organizations may instead be recognized formally or informally as arts districts (Chapple et al., 2010).

Similar to cultural districts, entertainment districts typically contain a number of big-ticket entertainment venues where culture and creative products are consumed, such as concert halls, opera houses, historic theatres and sports arenas. Entertainment zones (EZs), described as ‘concentrated nightlife districts occupying the margins of downtowns in former commercial and industrial areas’, tend to be less formal than entertainment districts, and cater to a decidedly younger crowd (Campo & Ryan, 2008, p. 292). They are generally characterized by older multi-story buildings that house an eclectic mix of modestly sized entertainment venues, including bars, restaurants, nightclubs and smaller music venues. At night, as club-goers and pub-crawlers begin to crowd the pavements, the EZ is said to exude a boisterous party-like atmosphere (Campo & Ryan, 2008).

Clusters of creative–cultural industries were evident among most CCDs, including those that have yet to be formally recognized as an arts/cultural/entertainment district. In all, 24 of the 102 CCDs carried an official special district designation, while 15 were classified as EZs because they were either identified as such by Campo & Ryan (2008), or exhibited typologies consistent with their conceptualization (see Appendix A). Agglomerations of creative and cultural amenities within CCDs, and other central city locations, are indicative of the strategic importance of centralized urban neighbourhoods in the creative–knowledge economy, despite advances in transportation and information technologies. Creative, cultural and knowledge industries are expected to benefit, as many industries do, from the cross-pollination and innovation catalysed by spatial proximity (Knudsen, Florida, Stolarick, & Gates, 2008; Mizzau & Montanari, 2008; Mommaas, 2004; Zheng, 2011). In addition to brick and mortar entertainment–cultural venues, CCDs often offer public spaces conducive to markets, festivals, concerts and other special events well-suited to the ‘ecstatic pleasures […] of music, color, feasting, and dance’ (Ehrenreich, 2006, p. 260).

While EZs are characterized by an eclectic mix of buildings that may feel old, or be perceived as old by their (mostly younger) patrons (Campo & Ryan, 2008), nearly all the neighbourhoods identified in this survey contained buildings listed on the National Register of Historic Places. Eighty-one CCDs were also associated with one or more historic districts implemented at the federal, state or local (county or municipal) level. Local historic districts typically offer the most robust protection of historical assets, but they also place the most severe restrictions on private property owners (Noonan & Krupka, 2011). Nevertheless, one of the most enduring characteristics of dynamic urban neighbourhoods is the impermanence of individual businesses and uses; 19th-century warehouses are converted into apartments and lofts, historic hotels are renovated and retrofitted for 21st-century offices, while old factories are converted into modern hotels. While the buildings and the basic footprint of the neighbourhood remain the same, the uses are constantly in flux, allowing older neighbourhoods to cater to current market trends and consumer preferences. In some CCDs, including the Old Market in Omaha, LoDo in Denver, and The Flats in Cleveland (see Appendix A), entire city districts comprised of former warehouses and industrial buildings have been converted into commercial and residential space, preserving the area’s industrial legacy while also attempting to accommodate the needs of the new creative–knowledge economy (Anderson & Winther, 2010; Florida, 2002; Rantisi, Leslie, & Christopherson, 2006; Zimmerman, 2008) (Figure 1).

Figure 1. Like many creative–cultural districts (CCDs), The Old Market in Omaha, Nebraska, is comprised primarily of repurposed warehouses and other formerly industrial spaces. The neighbourhood is a hotbed of creative–cultural production and consumption as evidenced by (A) artist studios and live–work units (pictured is the Bemis Center for Contemporary Arts), (B) pedestrian-oriented spaces and a lively café culture, (C) informal street festivals and markets, and (D) a variety of renovated urban lofts and apartments.

About half of the identified CCDs featured an extra layer of local governance known as a business improvement district (BID). Defined as ‘self-assessment districts that are usually initiated and governed by property or business owners, enabled by state laws, and authorized by local governments to provide public services in designated urban or suburban areas’, BIDs typically raise money either through taxation or special assessments (Morçöl & Wolf, 2010, p. 906). While BIDs vary considerably in size, structure and influence, most provide three basic services: security; maintenance/capital improvements; and promotion and marketing. To enhance security, BIDs have upgraded street lights, installed passive surveillance systems, and hired private security forces to patrol and discourage unwanted behaviour. Some have noted that this strict control of space is reminiscent of the suburban shopping mall (Mallett, 1994; Ward, 2007). Whereas the shopping mall is a private space masked as public (Goss, 1993), the formation of a BID ‘involves public space (parks, streets, and walkways) coming under the control of an institution formed by private interests’ (Ward, 2007, p. 784). To encourage middle-class consumption, BIDs not only need to feel safe, they must also appear well-maintained and cared for. The Providence, Rhode Island, Downtown Improvement District (DID), for example, employs a ‘Clean Team’ charged with sweeping pavements, collecting rubbish and litter, removing stickers and graffiti, painting street furniture, and shovelling snow in winter (Downtown Providence, 2013). BIDs, and other representative organizations, market themselves using a variety of media, including promotional websites. Cultural amenities, attractions and festivals are commonly featured on websites as part of a larger strategy to promote economic development and urban revitalization (Pratt, 2008; Tremblay & Battaglia, 2012). Marketing campaigns are often directed toward Florida’s (2002) creative class, and young professionals seeking ‘cool’ and ‘creative’ neighbourhoods with all the amenities needed to support their ‘Live, Work, Play’ lifestyles and ‘Shop, Eat, Drink’ consumption habits.

Additional place characteristics

Arguably one of the most appealing attributes of urban living is the convenience afforded by a walkable environment – an advantage that appears to be of particular value to members of Generation Y (Urban Land Institute, 2013). The key to walkability is proximity: the more amenities that exist within walking distance, the more likely people are to walk (Agrawal, Schlossberg, & Irivin, 2008). Traditional neighbourhood design – a feature common to CCDs, and pre-war neighbourhoods in general – involves the mixing of different land-use types such that residences are usually located within a few blocks of retail stores and other commercial areas (Cervero, 2003; Campoli, 2012). Design elements common among CCDs, including buildings with street frontage, wide and inviting pavements, street art and greenery, interesting signage and window dressings, street cafés, and a continuous facade, help create a well-defined, human-scaled urban environment that promotes pedestrian activity (Speck, 2012).

As human-scaled, pedestrian-oriented environments with a rich mix of uses, many CCDs provide a ‘sublime habitat’ for informal social spaces. CCDs typically contain several ‘third places’, described by Ray Oldenburg in The Great Good Place (1999) as ‘the typical gathering places of informal public life’ (p. 20). Examples of third places include bars/taverns/pubs, coffee houses, cafés, barber shops and beauty parlours, bookshops, and lounges. The third place provides a neutral space for social encounters that are independent of the associations at home (the ‘first place’) and at work (the ‘second place’). Most residents of CCDs need only walk a few blocks to reach the nearest third place. In suburban neighbourhoods, however, where residential and commercial uses are segregated and destinations are often scattered at significant distances and in different directions, the probability of a corner bar, coffee shop or other third place within a reasonable walking distance are comparatively low. In Bowling Alone, Robert Putnam posits that the decrease in civic engagement and social interaction among Americans in the late 20th century may have been contributable in part to post-war suburbanization and the formation of homogenous enclaves and gated communities that lack effective public spaces: ‘Gated communities are innately introverted, as traditional urban neighbourhoods were innately extroverted. Segregatory zoning policies have excluded such gathering places as local shops and restaurants from residential areas […]’ (Putnam, 2000, pp. 210–211).

The walkability of each CCD was estimated using walkscore.com, a free online service that provides a walk score for specific locations, neighbourhoods and cities based on the proximity of common amenities such as coffee shops, restaurants, grocery stores, drug stores, schools and bars. Appendix A lists the walk score for each CCD and the city in which it is located. The higher the walk score, the more walkable the area: walk scores above 90 are considered ‘a walker’s paradise’, those above 70 are ‘very walkable’, and those below 50 are classified as ‘car dependent’. Walks scores were obtained by first selecting a city, then indicating on an interactive map a location indicative of the centre of the CCD (usually a central intersection within the neighbourhood). Amenities within 1.5 km of the selected location were used by walkscore.com to calculate the area’s walk score. The average walk score for a CCD was 87.9, significantly (p < 0.01) higher than for cities as a whole, which scored 52.6. The desirability of pedestrian- and transit-oriented neighbourhoods is often evident in price premiums, which may be as high as 40–100% relative to more auto-dependent developments (Bartholomew & Ewing, 2011).

While public transit in the United States lags behind Europe and much of the Global North, popular support and investment in transit are on the rise (Department of Transportation, 2010). All 102 CCDs in the survey received some type of bus service; 19 also featured either rapid transit (subways, etc.) or light rail. Transit-oriented development (TOD) – particularly along rail lines – has become a popular approach to integrated transportation and land-use planning, though nearly all capital investments have been limited to metropolitan areas with populations of 1 million or more (Department of Transportation, 2010; Renne & Ewing, 2013).

Assessing neighbourhood change

For many US cities, the repopulation of urban neighbourhoods represents a chance to reduce the inequalities between city and suburb that arose in the 1960s and 1970s. Yet, as the old industrial economy gives way to the new knowledge economy, and the entrepreneurial city evolves in accordance with creative class aspirations for diversity, community and culture, cities and neighbourhoods are increasingly confronted with the exclusionary and homogenizing effects of gentrification (Kohn, 2013; Wyly & Hammel, 2004; Zukin, 2010). Zukin (2008, 2010) has been particularly critical of what Peck (2005) has described as ‘creative gentrification’ or ‘hipsterization’. She suggests that gentrifiers, yuppies and even cultural producers – young starving artists and struggling musicians among them – may displace existing residents not only by raising property values, but also by transforming spaces of consumption to suit their ‘bourgeois bohemian’ taste for lattes, organic produce and assorted boutiques (Brooks, 2000; Zukin, 2010). It is likely that low-income residents will at least pay more for housing (as much as 60% of their income) and other amenities (Freeman & Braconi, 2004). By altering neighbourhoods to fit their own needs and desires, gentrifiers may undermine the very cosmopolitanism and authentic character that attracted them (Douglas, 2012; Zukin, 2010). Some gentrifiers will actively avoid areas they deem too gentrified, at times purposely relocating to the leading edge of gentrification to escape the ‘fully established, too-gentrified-to-still-be-hip core’ (Douglas, 2012, p. 3586).

CCDs may provide an initial focal point for ‘urban pioneers’, who later move successively outward to colonize more affordable areas. Though this process has been well documented in North America’s largest cities, particularly in New York City (Lees, 2003; Newman & Wyly, 2006; Porter, 2010; Slater, 2004), there remains a need for empirical, neighbourhood-level data on cities further down the urban hierarchy. To begin addressing the paucity of empirical data on gentrification among mid-sized American cities, and to provide an initial exploratory assessment of recent neighbourhood change and gentrification among CCDs, several common indicators were examined, including population, household size, age, occupation, educational attainment (i.e., per cent of residents 25 years and older with a bachelor’s degree or higher), median household income, and median gross monthly rent (cost of the living space plus utilities). Data were obtained from the US Census Bureau’s American Community Survey (ACS) for each of the 102 CCDs at the scale of census tracts for the years 1999/2000 and 2010 (all data except population were five-year [2006–10] averages). Median household income and median gross monthly rent for the year 2000 were adjusted for inflation using 2008 as the reference year. While census tracts rarely align perfectly with neighbourhood boundaries, they do provide the most acceptable neighbourhood-scale approximation among the Census Bureau’s data products. Because census tracts are frequently combined or split between censes, it was first necessary to identify matching tracts in accordance with the census tract relationship file (US Census Bureau, 2010). A corresponding set of data for each indicator were collected for each of the 70 metropolitan areas to provide reference.

Demographic indicators

The data indicate that the majority of CCDs underwent significant demographic changes during the 2000s decade. Of the 102 CCDs identified in this survey, 80 gained population between 2000 and 2010, with an average increase of 23.4% (see Appendix A and Figure 2). For comparison, the population of the 70 metropolitan areas increased an average of 11% over the decade. Although the per cent increase in population for many CCDs between 2000 and 2010 outpaced their respective metropolitan areas, the numerical gain in population was often modest. The average numerical gain per CCD was 729, with an average initial population of 3108 in 2000 and an average final population of 3837 in 2010 (see Appendix A). Thus, given their limited initial populations, a 100% or more increase in population between 2000 and 2010 may have resulted simply from the construction or repurposing of one or two new residential structures. Nevertheless, the population growth of several CCDs is representative of a substantial influx of new residents. Among those with the highest numerical and per cent gains in population included the Historic Third Ward in Milwaukee, The Warehouse District in Minneapolis, The District in Nashville, and the Pearl District in Portland, Oregon. Altogether, the 102 CCDs gained just over 75,000 new residents, representing, on average, about 1% of total metropolitan population growth within the sample. This should not, however, be interpreted as the proportion of total metropolitan growth attributable to central city neighbourhoods in general; the CCDs identified here account for only fraction of all central urban neighbourhoods. Although population growth at the metropolitan level likely drove some of the population growth observed at the neighbourhood level, the two growth rates were not significantly correlated.

Figure 2. Key indicators of demographic and socio-economic change among CCDs (N = 102) and their respective metropolitan areas (N = 70) between 2000 and 2010.

CCDs with population gains were located throughout the country; those with losses were most prevalent in the eastern United States, particularly Florida. Pensacola, Naples, Fort Myers, Daytona Beach and Jacksonville, all in Florida, all contained CCDs that declined in population during the 2000s. With an economy that is based primarily on tourism and low-skill/low-pay service jobs, long-term unemployment rates that are among the highest in the nation, and a population that is considerably older than the national average, many cities in Florida are currently facing a ‘brain drain’ as educated young professionals seek better employment opportunities elsewhere (Acosta, 2011; Luhby, 2012; Rothwell, 2012; Stratton, 2013). Several cities in the ‘Rust Belt’ (e.g., Youngstown, Ohio; Flint, Michigan; and Erie, Pennsylvania) are also struggling to attract and retain creative–knowledge workers, particularly young college-educated professionals (Florida, 2010, 2012). Without the right economic climate, cities may have difficulty cultivating the social and diverse people climate, and thus also the vibrant, mixed-use urban neighbourhoods, that many young professionals desire (Dewan, 2006; Florida, 2002; Zenker, 2009). Florida (2002, 2008) has argued that the opposite may be true, as well: cities that fail to attract young professionals and members of the creative class will be less economically competitive. Though contested (Boschma & Fritsch, 2009; Donegan, Drucker, Goldstein, Lowe, & Malizia, 2008), this logic has fuelled many of the latest rebranding and promotional strategies (Long, 2009; Zimmerman, 2008).

An analysis of variance (ANOVA) was performed to compare the population growth among CCDs with and without historic districts, rail access, BIDs or arts/cultural/entertainment district designations to determine if any of these features may have played a significant role in promoting development. The only factor that was statistically significant (p = 0.026) at the 5% level was presence/absence of a rail connection. The 19 CCDs with rail access experienced an average population increase of 66%, while the remaining 83 CCDs without rail access saw an average population increase of only 29%. Proximity to rail therefore stands out as one of the most probable effective catalysts for growth among neighbourhoods in mid-sized and large US metropolitan areas.

A brief comparison of household size and the prevalence of two age cohorts (young: 20–34 years; seniors: 62+ years) among CCDs and their respective metropolitan areas yielded support for the notion that CCDs have been particularly attractive to young singles and couples without children. The data also suggest, however, that many CCDs may not be the young neo-bohemian enclaves they once were. Among CCDs, 20–34 year olds represented 53% of the population in 2000 (compared with 21% at the metropolitan scale); by 2010 this proportion had fallen to 33%, a loss of 36,000 young adults (Figure 2). The increase in population experienced by most CCDs, therefore, was not driven by the young; nor was it driven primarily by an influx in older (62+) residents, whose numbers increased a modest 8%. As a proportion of all CCD residents, in fact, those aged 62+ declined from 13.8% to 12.0%. Evidently, CCDs became increasingly attractive to those in their middle years (i.e., 35–61 years old). Yet, it is interesting that the average household size among CCDs (mean = 1.65), did not vary significantly between 2000 and 2010, suggesting that family structure – particularly the enduring pattern of fewer and smaller families within central urban neighbourhoods – did not change appreciably over the decade. Clearly, there still exists a strong life cycle element, with families generally preferring suburban locations with child-oriented amenities such as large yards and parks, shopping centres, quiet residential subdivisions, and high-achieving schools (Frenkel, Bendit, & Kaplan, 2013; Lawton et al., 2013; van Oort, Weterings, & Verlinde, 2003). However, as more people live alone and have fewer children (US Census Bureau, 2013), the convenience of proximity – especially to work and to entertainment – inherent in urban environments is likely to become increasingly advantageous for those in their middle years and at the peak of their professional careers. The fact that CCDs are losing younger residents may be a symptom of this transformation. Younger residents, many of whom are at the start of their careers, may be leaving CCDs in search of more affordable, if not ‘hipper’, neighbourhoods. Another probable factor at play is the impact of the Great Recession (2007–09) during the latter years of the decade. Using data from the US Census Bureau, Mykyta (2012) concluded that the recession years saw a rise in young adults aged 18–34 with ‘dependent living arrangements’, as unemployed and under-employed high school and college graduates either returned home or never left.

CCDs that were successful at retaining younger populations also tended to experience the most significant gains in educational attainment (r = 0.439; p < 0.001; Figure 3). One potential explanation is that CCDs were generally better at attracting and retaining those college-educated young professionals who could afford to weather the economic recession in-place, or, in some cases, even migrate to particularly attractive urban neighbourhoods exhibiting gentrification and rising rents. It is worth noting that only five CCDs not only saw growth in the absolute number of young people, but also in the proportion of residents aged 20–34: LoDo in Denver, The Strip District in Pittsburgh, Little Italy in San Diego, Downtown West in St. Louis, and Ybor City in Tampa. Each of these four CCDs experienced significant increases in overall population, income, rent, education and the proportion of workers employed in high-skill ‘creative’ occupations (described below) between 2000 and 2006–10.

Figure 3. For 102 CCDs in 70 mid-sized metropolitan areas, the relationship between the change (2000–10) in the proportion of residents aged 20–34 years and the proportion of civilian workers with a bachelor’s degree or higher.

Economic indicators

As nodes of creative–cultural activity, CCDs were expected to exhibit a high proportion of workers with creative, knowledge and cultural occupations (McGranahan & Wojan, 2007). The US Census Bureau’s occupational categories ‘management, professional, and related occupations’ (2000) and ‘management, business, science, and arts occupations’ (2010) were used to approximate the number of highly skilled workers in creative fields at the census tract level. These are broad occupational categories and estimations only; many ‘creative’ jobs are likely to fall outside these categories while some non-creative jobs are likely to exist within them. Nonetheless, they are expected to provide a reasonable proxy of ‘creative class’ employment at the sub-county scale (Florida, 2013b).

As anticipated, CCDs boasted a significantly (p < 0.001) higher proportion of workers in creative–knowledge occupations relative to metropolitan areas, both in 2000 and in 2006–10. Over the decade, the average creative class share within CCDs grew from 40% to 48% (Figure 2). Metropolitan areas saw a more modest increase in the proportion of creative knowledge workers, which grew from 34.3% to 36.0%. The metropolitan-level estimates align well with those calculated by Florida (2012), who used more detailed occupational data from the Bureau of Labor Statistics’ (BLS) Occupational Employment Survey (OES). (Note: OES data are not currently available for census tracts; the US Census Bureau ACS data was therefore used at both the census tract and metropolitan scale for consistency.) Thus, while many CCDs lost young people during the 2000s, the vast majority (77 of 102) gained highly skilled workers. Because creative–knowledge jobs generally require college degrees, and pay higher wages than those in either the working- or service-class occupations, it is not unexpected that change in per cent creative class share was significantly and positively correlated with change in education (r = 0.543, p < 0.001) and income (r = 0.355, p < 0.001). Among mid-sized US metropolitan areas, as with those at the top of the nation’s urban hierarchy (Florida, 2013; Howell, 2005; Lloyd, 2002), central urban neighbourhoods with the right mix of physical, cultural and social/entertainment amenities are quickly evolving into creative-class enclaves.

Average increases in educational attainment, income and gross monthly rent among CCDs also significantly outpaced those of their respective metropolitan areas (see Appendix A and Figure 2). Educational attainment, as measured by the per cent of residents with a bachelor’s degree or higher, rose from 32% to 43%, on average, among CCDs. The 70 metropolitan areas, by contrast, saw a more modest increase from 26% to 29%. Note that CCDs not only experienced a more rapid increase in education, on average, they also exhibited higher than average educational attainment within their respective metropolitan areas. In fact, the educational gap only widened between 2000 and 2008. The relatively high degree of educational attainment, and rise in educational attainment, among many CCDs not only suggests gentrification, it also supports the notion that these neighbourhoods are increasingly attractive to well-educated workers. This appears to have been particularly true among CCDs in close proximity to a rail line. An ANOVA indicated that, like population growth, those CCDs with rail access experienced a significantly (p = 0.027) steeper rise in the proportion of residents with a four-year college degree than those with bus service only.

While CCDs are clearly a magnet for human capital as measured by educational attainment, analyses using occupation-based measures (including the examination herein of workers with creative–knowledge occupations) have likewise indicated that highly skilled workers are particularly attracted to central urban neighbourhoods, as well as low-density suburban and exurban areas, depending on life style preferences and life cycle stage (Florida, 2013; Frenkel et al., 2013; Lawton et al., 2013; McGranahan & Wojan, 2007). This polarization of human and creative capital in gentrified older neighbourhoods and new suburban enclaves in metropolitan America has paralleled a rise in income inequality and residential segregation by income in recent decades (Fry & Taylor, 2012; Watson, 2009).

As further evidence of gentrification, median household income in CCDs rose significantly (p < 0.001) during the 2000s, increasing 19% from an average inflation adjusted US$33,832 in 1999 to US$40,054 averaged over 2006–10 (Figure 2). This increase occurred in spite of the Great Recession of 2007–09, during which time inflation-adjusted median household income declined 4% nationwide (Kochar, 2012). As a whole, the 70 metropolitan areas did not fare as well as CCDs: median household incomes declined 7.1% from US$55,101 in 1999 to US$51,167 in 2006–10. Note that the average median household incomes of CCDs were generally below those of their respective metropolitan areas; this was likely due to smaller household sizes and proportionally fewer dual income-households in these neighbourhoods. Per cent change in education and median household income, though significantly correlated (r = 0.525, p < 0.001), exhibited minimal regional specificity; CCDs with the most and least substantial gains in either dimension were spread geographically throughout the country. There were also no significant associations between changes in income among the sample of CCDs and walk score, or the presence or absence of historic districts, BIDs, arts/cultural/entertainment districts, or rail-based transit. To assess accurately the potential impacts of special districts on neighbourhood change, however, it would be necessary to compare indicators before and after district implementation, as well as to control for confounding factors. Macdonald, Golinelli, Stokes, & Bluthenthal (2010), for example, compared crime rates among neighbourhoods in Los Angeles before and after BID formation, and concluded that the implementation of BIDs was associated with a significant decline in robbery and violent crimes.

On average, CCDs experienced a 28% increase in inflation-adjusted median gross monthly rent between 2000 and 2006–10 (Figure 2). Their respective metropolitan areas, by comparison, saw only an 8.6% average increase. In 2000, the average residential unit in CCDs rented for US$643, substantially less than the US$728 average rent at the metropolitan scale. By the end of the decade, however, the average rent among CCDs had risen to US$821, surpassing the metropolitan average of US$791. The increase in gross monthly rent among CCDs was significantly correlated with both the rise in educational attainment (r = 0.408, p < 0.001) and the rise in median household income (r = 0.435, p < 0.001). However, change in gross monthly rent was not associated with the presence of special districts, transit or walk score. It is not surprising that CCDs with higher walk scores were not associated with higher incomes, rents or education levels; the difference between the highest and lowest walk scores was minimal (i.e., both the low score of 66 and the high score of 100 generally indicate walkable environments), and the presence of any number of other amenities may have affected demand.

Particularly revealing was the change in gross monthly rent relative to the change in median household income. Building on empirical work by Roback (1982), Gyourko & Tracy (1989), and Glaeser, Kolko, & Saiz (2001), Glaeser & Gottlieb (2006) concluded that housing prices rise faster than nominal wages (as opposed to ‘real wages’ that take into account cost of living), primarily in response to an increase in the number or quality of urban amenities. Natural amenities such as bodies of water can also influence the cost ratio of housing to wages; however, they are much less likely to change over short time periods. Although Glaeser & Gottlieb (2006) performed their analyses at the city scale, it is reasonable to extend their logic to neighbourhoods. An important caveat, however, is that inferences made at this finer scale are likely less powerful given that demand for housing is necessarily affected by conditions in surrounding neighbourhoods as well as the encompassing metropolitan area. With that said, it appears that urban amenities were on the rise in the majority of CCDs: in 59 of the 102 neighbourhoods, the per cent increase in gross monthly rent exceeded per cent gains in gross monthly income (Figure 4). Note that 21 of the 59 neighbourhoods actually experienced a reduction in inflation-adjusted income in step with the economic downturn. Nationwide, lower income neighbourhoods suffered larger per cent declines in wealth during the Great Recession (Lerman & Zhang, 2012), a pattern that is also reflected in this sample of neighbourhoods. Rent continued to increase in many CCDs however, suggesting that demand for housing at the neighbourhood and/or metropolitan scale grew or remained stable. Among those CCDs with both positive gains in rent and income, LoDo in Denver, the Strip District in Pittsburgh, the Lower Garden District in New Orleans, and Pike Place Market in Seattle experienced the greatest gain in gross monthly rent relative to median household income. Not only are these neighbourhoods becoming wealthier; the demand for housing units likely exceeds supply, contributing to the rising cost. The eventual displacement of lower income groups (including artists and other members of Florida’s ‘creative core’) is among the anticipated consequences of such transformations; a common point of criticism levied at creative city policies (Atkinson & Easthope, 2009; Catungal et al., 2009; McCann, 2007). The need to provide affordable housing, however, is readily acknowledged by Florida (2013a), who suggests that a ‘new social compact’ is required to both increase wages among those in service and working-class occupations, and make ‘housing more affordable by increasing density, and making urban centers more accessible by improving transit’.

Figure 4. For 102 CCDs in 70 mid-sized metropolitan areas, the relationship between the per cent change (2000–10) in gross monthly rent and the per cent change in median household income. CCDs left of the dotted line likely experienced an increase in urban amenities.

In 29 additional CCDs, both income and rent increased, but the rise in income exceeded the rise in rent. The most extreme examples included Covington, Kentucky, in Cincinnati, Downtown West in St. Louis, and City Market in Roanoke. In these neighbourhoods, housing became increasingly affordable to an increasingly wealthy group of incoming residents. The desirability of these neighbourhoods is clearly on the rise, perhaps more due to their relative affordability and strategic central locations rather than growth in amenities. Eight CCDs experienced both a reduction in income and in rent, suggesting a decrease in demand potentially due to a reduction in amenities, or, more likely, an overabundance of housing in combination with a downturn in the local economy (Kuehn, 2011). Only four CCDs (Old World 3rd Street, Milwaukee; The Flats, Cleveland; Glenwood South, Raleigh; and LoDa, Mobile) saw a decline in rent and an increase in income, suggesting that housing supply outpaced demand despite rising incomes.

Finally, a two-step cluster analysis suggested that the 102 CCDs could be logically organized into three fundamental categories based on the economic variables explored above: (1) those with relatively affluent populations in both 2000 and 2010, and exhibited high incomes, education, rents, and a large proportion of workers in creative–knowledge occupations, but did not change significantly in any of these areas over the time period (Cluster 1; N = 42); (2) those that were less affluent in 2000 and remained so over the decade, with relatively low incomes and education (Cluster 2; N = 42); and (3) those most likely to have experienced significant gentrification, as demonstrated by rising incomes, rents, education, and the proportion of workers employed in creative–knowledge occupations (Cluster 3; N = 17) (see Appendix A and Figures 3 and 4). Although CCDs in the second and third cluster exhibited similar metrics in 2000, the third cluster was essentially indistinguishable from the first cluster by 2006–10.

CCDs in each of the three clusters face unique challenges and opportunities moving forward. Neighbourhoods in the second cluster will likely be heavily growth-oriented in the coming years, with marketing campaigns and public–private partnerships aimed at luring any development that seems within reach. There is a unique opportunity for associated local governments to incorporate inclusivity and equitability into their long-range plans. CCDs in the first, and especially the third, cluster will look to expand their influence – if not their physical footprints – into surrounding neighbourhoods. In future work, the social and spatial evolution of CCDs may be traced using a more complete assortment of socio-economic and demographic indicators. Important questions remain as to the potential for gentrification within CCDs to catalyse neighbourhood change in surrounding urban communities, and whether the extent and nature of these changes may be similar among the nation’s regional (e.g., St. Louis, Portland, Charlotte) and global (e.g., New York, Chicago, Los Angeles) urban centres. Furthermore, indices of diversity could be applied to this, or an expanded, set of neighbourhoods to address outstanding questions regarding the impact of gentrification on neighbourhood-level social and cultural diversity (Freeman, 2009).

Conclusions

In the early 21st century, vibrant bohemian neighbourhoods have become an increasingly attractive alternative to the ubiquitous auto-centric suburban lifestyle in the United States, particularly to the affluent and well-educated. Although creative–knowledge workers have shown an affinity for a variety of neighbourhood typologies, it is evident that the latest wave of urban revitalization and gentrification has been propelled in part by the propensity of some within this cohort to seek out uniquely urban amenities, environments and life styles. This may be particularly true of singles and couples without children for whom suburban living may yield fewer advantages.

The preliminary analysis presented here of the demographic and economic profiles of 102 CCDs in 70 mid-sized US metropolitan areas yielded important insights into the nature and magnitude of recent (2000–10) neighbourhood change. The majority of the CCDs in the survey experienced population growth, which exceeded that of their metropolitan regions, while also exhibiting signs of gentrification, most notably rising incomes, rents, education, and the proportion of workers employed in creative–knowledge fields. However, there was no significant (p < 0.05) difference in the change in population, income or education levels among CCDs that had or had not implemented BIDs and other special districts (i.e., arts, cultural, entertainment and historical), suggesting that such designations may have had limited effect on neighbourhood change over this time period. Research that examines an expanded set of variables over a longer time frame and specifically compares pre- and post-designation conditions is needed to more thoroughly address this question. Lastly, one of the most intriguing and unexpected results was the near-universal reduction in the proportion of young adults in the much sought-after 20–34-year-old demographic. CCDs appear to be losing their identity as young neo-bohemian enclaves. In recent years, many have transformed into fully gentrified cultural–entertainment machines dominated by the ‘creative class’ or the ‘educated elite’. Among a sizable cohort of CCDs, not even the Great Recession of 2007–09 could dramatically slow this transformation.

For central urban districts that have struggled for decades to remain viable and relevant in the face of suburbanization, edge cities, and high-speed electronic communications, the mere possibility of urban resurgence through creative–cultural consumption and production remains undeniably alluring. Such transformations, however, are rarely accompanied by enhanced equitability. Even Florida (2013b) has admitted that the benefits of a growing creative–knowledge economy ‘accrue mainly to knowledge, professional, and creative workers’ whose wages generally rise faster than expenses. For members of the working and service classes (which typically comprise 60–70% of the workforce), however, rising wages rarely keep pace with expenses. There is great risk, then, in creative–cultural neighbourhoods evolving ever further into homogenous and exclusionary upper-middle-class enclaves, or little more than playgrounds for the burgeoning ‘creative elite’. If anything, communities and municipalities have proven largely ill-equipped to maintain and nurture equitability in the face of gentrification, particularly when so many downtowns and central urban neighbourhoods continue to struggle, and the competition to attract skilled workers both between and within metropolitan regions is fierce. CCDs are therefore prime battlegrounds for the equitability-versus-growth debate, and their fate over the coming years and decades are likely to shape urban geographies far more than the changes in their absolute numbers might suggest.

Appendix A

Table A1. Select attributes and recent neighbourhood change (2000–10) among ‘creative–cultural districts’ (CCDs) in mid-sized US metropolitan areas.