Yaoyi Zhou

DEA, Cornell University (yz774@cornell.edu)

Abstract

Coworking spaces are currently fast-growing in the cities, and it represents a new living and working lifestyle for an increasing population of contingent workers. However, what kind of urban environment are the coworking space embedded in? Few studies have analyzed the relationships between the emerging coworking workplace and the urban environment. This study is an exploratory research for this emerging urban ‘lifestyle,’ which aims at developing an understanding of the coworking spaces’ locations and the urban context so that additional questions or hypotheses about the meaning of coworking as a sustainable urban development can be constructed. The location analysis answers the question about where the coworking spaces locate and what the factors are correlated with its clustering pattern. Locations of 123 coworking spaces in Manhattan, NYC are analyzed through Geographical Information System (GIS) platform ArcGIS in this study. This analysis is based on a framework of factors in five categories: 1) Neighborhood and social atmosphere; 2) Neighborhood environment; 3) Transportation connection; 4) Random discovery and events; 5) Life convenience. Eighteen factors including population density, median household income, proximity to subway stations, accessibility to green spaces and so on are examined in the location analysis, ordinary least squares regression analysis, and geographically weighted regression analysis. Most significant correlations are found between the coworking spaces location and transportation connection, random discovery and events, and life convenience. The result suggests that the coworking spaces on the market are not randomly distributed but highly clustered, and they are mostly located in the mixed-use areas so that it allows easier access to the surrounding facilities and urban resources. Planners, workplace designers, and potential investors will benefit from the study result, as the cluster relationships between coworking spaces and the social, environmental factors are revealed from the maps and analysis.

Keywords: Coworking; Workplace; Urban environment; Lifestyle; Location Analysis; GIS.

Introduction

The urban environment is changing as more and more millennials choose to live and work in the cities. The growing number of coworking spaces echoes with the concept of sharing economy and indicates an emerging urban lifestyle for millennials and an increasing population of contingent workers. It advocates a self-directed, collaborative and flexible workstyle which is based on sharing and mutual trust. The coworking industry made up 0.7 percent of the total U.S. office market in 2017, but demand is unprecedented and fueling the growth of large providers in major markets (JLL, 2017). According to Deskmag (2018), an online journal for coworking’s estimate, more than 1.7million people will be working in around 19,000 coworking spaces around the world by the end of 2018. The trend of expansion is continuously strong (Deskmag, 2018). New York City is the most prosperous market of coworking spaces (Moriset, 2014). For the first eight months of the year 2018, coworking space companies leased a total of 1.9 M SF in Manhattan, NYC, which accounts for nearly 10% of all new leases in 2018 (Hall, 2018). According to data retrieved from Coworker, the largest online coworking space search website and coworker community, there are 176 coworking spaces in the great New York City area at Nov. 2018.

However, where are the coworking spaces? Few studies have explored coworking spaces’ locations and what kind of social and environmental factors should planners or designers care about to achieve successful coworking space development. Taking Manhattan, NYC as a study case, this study explores the relationship between coworking spaces and the urban environment they embedded in, to answer the question how should we understand coworking as an emerging urban lifestyle.

Millennials’ Urban Lifestyle

Life was good for Millennials until they enter the stagnant job market. Given their educational credentials, Millennial had every reason to be optimistic about their future career. However, the job market Millennials face is tough. A large portion of well-paid employment disappeared due to increased outsourcing practices, and the employers are more relying on flexible work arrangements, which makes permanent full-time employment less likely to happen (Tolbert, 1998; Banerjee, Tolbert & DiCiccio, 2012). Millennials report disproportional unemployment rates and often find themselves under-employed (The Economist, 2016). They are more likely to encounter a non-standard work arrangement. This kind of work can be mostly viewed as flexible, short-term, contract-based, and is referred to as part of the "gig economy." In the US, more than 80% of employment growth between 2005 to 2013 involved workers under alternative contract arrangements rather than traditional ones (Friedman, 2014).

Beside of the competitive job market, Millennials also confront to expensive housing markets. The high housing price is also a result of the tendency for jobs, especially the type of employment appealing to well-educated Millennials, concentrating in large cities. As a result of difficult employment and housing conditions, a higher proportion of Millennials extend the young adult phase of their lives and delay family and child-rearing life stage. Watters (2003) sees the importance of networks of friends for Millennials, and they are more appealed to urban living. Multiple reasons explain why millennials ditch suburban and choose an urban lifestyle, such as a decrease in driving and decrease employment (McDonald, 2015; Hot in the City, 2014).

The concentration of young adults in the central urban area has increased dramatically since the ‘urban renaissance' starting from as early as the 1960s in North America (Atkinson, 2004). Moos (2016) describes the process that planners and policymakers use those urban amenities to attract young knowledge workers as ‘youthification,’ which occurs as young adults increase in the share of the total population in specific neighborhoods. The urban image Millennials evoke that young adult living in attractive, edgy inner-city neighborhoods filled with café, bars, shops, gyms and providing active music scenes. The amenities emphasized in the innovation-oriented, talent-attraction strategies of many US and Canadian cities catered to the preferences of well-educated, affluent Millennials (Moos, Pfeiffer & Vinodrai, 2017). Coworking spaces, innovation districts, makers-spaces previously have tried different ways to use the physical design of spaces to match with the new urban lifestyle and promote next urban development. In recent years, coworking space is an emerging urban phenomenon and an increasing staple of real estate deals in large cities in North America. It is also becoming a symbol of authentic urban living experience in the knowledge economy.

Contingent Workers and Coworking Spaces

Coworking spaces are shared workplaces occupied by different sorts of knowledge professionals, mostly freelancers, working in various degrees of specialization in the vast domain of the knowledge industry (Gandini, 2015). The development of coworking spaces roots in the changes in workforce demographics, and the way people work and get hired. The emergence of a considerable number of contingent workers and contracting professionals start from the changes in organizations from the 1980s (Barley & Kunda, 2004). According to a report that the US Government Accountability Office published in 2015, the size of the contingent workforce can range from less than 5% to more than a third of the total employed labor force which is about 5 million in 2010, depending on the definition of contingent work and the data source (Jeszeck, 2015).  The estimation of contingent workers population is difficult, and varies dramatically because of the different definition of ‘contingent workers.’ A much larger number, 53 million, which Freelancer Union published in 2012 and 57.6 million in 2017 are also another cited source for self-employed workers population (Union & Desk, 2012; Union & Upwork, 2017). Responding to the growth of contingent workers, coworking was described as a bottom-up solution or collective strategy for coping with structural changes in the changing labor market, especially in creative industries (Markel, 2015). Spinuzzi (2012) studied coworking spaces in Austin (Texas) and suggested that coworking is the new model of ‘distributed work’ which respond to incoming trend in the organization of labor in the knowledge economy.

Neumann, the founder of WeWork, the largest coworking space provider in the world, said WeWork is neither a real-estate company nor a tech company. “We Generation, craves sharing and collaboration rather than isolated offices. They’re coming to us for energy, for culture.” Neumann described WeWork as a lifestyle or community-focused company, according to people familiar with the instructions (Eliot, 2017). The coworking spaces accommodate work practices that are typical for mobile, project-based work which could be carried out ‘anywhere,’ but more importantly, it celebrates the idea of mutual support among freelancers and self-employed workers (Merkel, 2015). It is a superclass that encompasses the good-neighbors and good-partners configurations as well as other possible configurations that similarly attempt to network activities within a given space (Spinuzzi, 2012). Moriset (2014) studied an overall number of 2,498 mapped spaces worldwide and found that coworking spaces clustered in “creative cities" and roots in the strongly connected local communities. It was interpreted as ‘third places’ between home and work, and its development aligns with policies which point towards the emergence of “creative districts” (Moriset, 2014). Different from telework centers which mostly located in rural and suburban areas and with a low degree of professional interaction (Moriset, 2010), coworking spaces emerged from the central urban area and celebrated the culture of resource sharing and the development of professionals' community.

However, previous research has ignored to explore the location of coworking spaces and has not examined the characteristics of the urban environment that they embedded in. While the prevalence of coworking spaces accommodating knowledge workers in large urban centers is well established, there is no study about their locations from a macro urban perspective. Analyzing the location of coworking spaces is timely, as the growth of the post-industrial, knowledge-intensive economy is continuously changing how the way people work, live and experience cities. It could shed light on this urban phenomenon, and it is informative for knowing more about the millennials and contingent workers’ lifestyle just by looking at they choose to work.

The Goal of This Study

This study is an exploratory research, which aims at developing an understanding of the location of coworking spaces, to explore how coworking spaces are clustered. The location analysis aims at examining the relationships between different variables in demographics, urban environment and the locations of the newly emerged coworking spaces. The factors that correlate the most with the location has potential in constructing for a location suitability model in the future. Planners, coworking space developers and managers will benefit from the study result of this project, as the cluster relationships between coworking spaces’ locations and social, environmental factors can be revealed from the maps and analysis.

Methods

Data and Sources

Location decisions are affected by a variety of determinants, which makes the process complicted, dynamic and difficult (Rymarzak & Siemińska, 2012). In this study, the list of factors is abstracted from the result of an existing coworking space market study survey (Deskmag, 2017a). Deskmag is an online magazine about coworking, its people and spaces (Deskmag, 2017b). Reflected from the result (as Shown in Figure 1), there were multiple location-related factors mentioned in the reasons for the question ‘why coworking spaces are chosen’. There were 59% of the participants chose a social & enjoyable atmosphere; 51% of them chose a close distance to home; 41% of them chose good transportation connection; around 30% chose events and random discoveries; 14% chose close to supermarkets and restaurants.

Figure 1: Reason why people choose Coworking spaces (Deskmag, 2017a).

According to the information retrieved from the survey result, the location-related factors are summarized in five factors as follows:

1) Neighborhood and social atmosphere

2) Neighborhood environment

3) Transportation connection

4) Random discovery and events

5) Life convenience

Instead of taking the whole New York City as the area to study, this study takes a closer look at the Manhattan area because it houses the majority of the existing coworking spaces. It is easy to tell that the existing location of coworking spaces are clustered in the midtown and downtown area in Manhattan (as shown in Map #1). As the goal of this research is to propose a model for understanding and evaluating what kind of location would be suitable for new coworking spaces, Census block group was chosen as the level of analysis for this project instead of Census tract. The main reason for not choosing the Census tract is that the Census tract will aggregate the counts of coworking space locations, and the result is not detailed enough to decide which area should take a closer look at. The factors are operationalized in a structure of variables, and they are described in Table 1.

Table 1 Data, types and sources

Majority of the data was already available online when it was retrieved at Dec. 2017. The demographic data such as population density, age, race job is obtained from the New York City Census FactFinder (2010-14 ACS). Point data including coffee shops and fitness centers and so on was acquired through SimpleAnalytics. The location data of restaurants was obtained from an existing list on the website of the NYC Restaurants database. The coworking spaces location data was downloaded as an excel file from a website of New worker magazine. There are 125 locations on the lists, and it only has the name and address. The coordinate of each coworking space was searched in google map by the author, and the price was recorded from each coworking space's website. Every coworking space on the list was checked through its website to see if it was still operating, and what price they offer for different desk and office choices. After the location verification process, there were 123 coworking spaces on the list, 39 of them are WeWork branch coworking spaces. In general, the coworking spaces provide four kinds of membership options, which includes Virtual tenant which only helps to receive mails; Hotdesk (Flexible desk); Assigned desk (Fixed desk in the open office); and Private office. Pricewise, the private office charges the highest price at a median number of $900/month, while Fixed desk for $600/month, Hotdesk for $375/month, and virtual desk for $50/month at a median number. The price varies among different coworking spaces and locations, but the majority of the coworking space provide price information for the private office, and it was taken as the price indicator as shown in Map#1.

For the polygon data of the block groups, blocks with empty data (population) were taken out during data preparation, along with the central park block which is an obvious outlier when calculating area and population. Finally, there are 1,084 block groups on the list for analysis. Eight factors with the highest correlation are shown in the following analysis section along with the maps generated by the author using ArcGIS. The rest of the exploratory analysis of the other ten factors are documented in Table 2.

Analysis and Results

Map #1: Coworking spaces location and Price. Map #1 shows the location of the 123 coworking spaces which are clustered in the Midtown and Downtown area. A few of them are scarcely distributed above the 60^th street (south end of the central park). Taking the price of the private office into consideration, the darkest blue dots (most expensive) are clustered in the Midtown area, and they overlay with the highest coworking space density (blocks in red has the largest number of coworking spaces). This map shows that coworking spaces are not randomly distributed in Manhattan, and there is a tendency that the coworking space with a higher price is clustered with other high-priced coworking spaces in several specific areas.

Map #1 Coworking spaces location and Price Map #2 Population density

Map #2: Population density. For map #2, the initial assumption is that there is a positive correlation between population density and coworking space. However, when population density and location density information is overlaid, it shows a negative correlation such that the coworking space concentrated area has lower population density. The data source could help to explain the result. One explanation for this phenomenon is that the population density data from the American Community Survey was measured mainly in the residential area. The population density in the map shows the residential density, which is influenced by land-use zoning regulation. As coworking space is in office use, it may cluster at the office and mixed-use zones which has lower residential population density.

Map #3: Household median income. As the global coworking survey result indicated that 59% of the members chose coworking space because of good social atmosphere, median household income is chosen as an indicator which represents the construct of social atmosphere. The Map above shows a relatively positive relationship between the number of coworking spaces and median household income. Ordinary Least Square Regression (OLSR) analysis for the two variables was conducted, taking the number of coworking spaces as the dependent variable and median income as the independent variable. The p-value turns out to be 0.000017 < 0.01. There is a significant statistical correlation between these two variables. Other demographic variables are also tested and showed in Table 2.

Map #3 Household median income Map #4 Housing median rent

Map #4: Housing median rent. As rent is usually an indicator of housing condition and the location of the property, housing median rent is chosen as an indicator for neighborhood environment. The Map above shows a relatively positive relationship between the number of coworking spaces and median house rent. In the result of Ordinary Least Square Regression (OLSR) analysis, taking the number of coworking as the dependent variable and median income as the independent variable, the p-value turns out to be 0.000005 < 0.01. There is a significant statistical correlation between these two variables.

Map #5: Subway stations. The relationship between public transportation and coworking location spaces is examined in this map. It is noticeable that the distance between two subway stations is smaller in the midtown and downtown area, and one block sometimes has multiple subway stations. In this case, Buffer function instead of Near function is used in ArcGIS to operationalized the accessibility of subway stations. The indicator is calculated by the number of subway stations in a buffered 0.1mile area of the block, divided by the area (0.01Acres) of the block. The analysis shows a significant positive relationship between the two variables. In OLSR, the p-value is 0.000064 < 0.01.

Map #5 Subway stations Map #6 Museums

Map #6: Museums. The number of Museums and Theaters in a block group is used to operationalize the chance of random discovery and events. The map below shows a positive relationship between the number of coworking spaces and the number of museums in the same block group. In the result of Ordinary Least Square Regression (OLSR) analysis, taking the number of coworking spaces as dependent variable and number of museums as the independent variable, the p-value turns out to be 0.000000 < 0.01. There is a statistical correlation between these two variables. However, the events may not be the best reason to explain the correlation between museums and coworking spaces. A further literature review of the location analysis of museums is needed.

Map #7 Grocery Map #8 Coffee shops

Map #7: Grocery. Numbers of groceries are used to operationalize life convenience, as it indicates accessibility to buy daily required items, and it is one of the reason in the result of the global coworking space survey. The map above shows a strong positive relationship between the number of coworking spaces and the number of groceries in the same block group. The map above shows a positive relationship between the number of coworking spaces and the number of groceries. In the result of Ordinary Least Square Regression (OLSR) analysis, the p-value turns out to be 0.000000 < 0.01. The correlation between these two variables is statistically significant.

Map #8: Coffee shops. As another indicator of life convenience, the number of coffee shops is tested to see its relationship with coworking space. The map above shows a strong positive relationship between the number of coworking spaces and the number of coffee shops in the same block group. The map above shows a positive relationship between the number of coworking spaces and the number of groceries. In the result of Ordinary Least Square Regression (OLSR) analysis, the p-value turns out to be 0.000000 < 0.01. The correlation between this two variables is statistically significant. The relationship between the two kinds of variables may be an indicator of the same group of market consumers. A further literature review of the factors determines the location of coffee shops will be helpful.

Table 2

Summary of the regression analysis result

* An asterisk next to a number indicates a statistically significant p-value (p <0.01)

Note: (OLSR): Ordinary Least Square Regression.

(GWR): Geographically Weighted Regression. 

As most of the factors in the above maps have shown a positive relationship with the coworking spaces' density in the block groups, which factor has the strongest correlation with the coworking space is the following question. To answer this question, Ordinary Least Square Regression analysis is conducted for each of the 17 factors, and the result is summarized in Table 2. Geographically Weighted Regression is also conducted, to see the correlation taking account of the geographical clustering condition. R-square data in the Geographically Weighted Regression are also summarized in Table 2. As the R-Square is a measure of goodness of fit, its value varies from 0.0 to 1.0, with higher values being preferable. It could also be interpreted as the proportion of dependent variable variance accounted for by the regression model. By looking at the Probability and R-Square number, the most correlated factor is identified. The factors with high statistical correlation are bolded in the table.

Table 3 The final list of 6 factors for predicting model in OLSR

* An asterisk next to a number indicates a statistically significant p-value (p <0.01)

Note: (GWR): Geographically Weighted Regression;  

(OLSR): Ordinary Least Square Regression

Table 4 The final list of 4 factors for predicting model in OLSR

* An asterisk next to a number indicates a statistically significant p-value (p <0.01)

Note: (GWR): Geographically Weighted Regression;

(OLSR): Ordinary Least Square Regression

Six factors with Probability smaller than 0.1, and R-Square large than 0.1 are identified and are grouped to run a new Ordinary Least Square Regression analysis with multiple factors. A new set of Probability value is calculated, and it turns out that only four of them (Subway station, Museum, Grocery, Coffee shops) has Probability value smaller than 0.1 (As shown in Table 3). In this case, a new model which is consisted of the remaining four factors is constructed, and the OLSR analysis result is present in Table 4. Two maps are also constructed to show the changes of Std. Residuals.

Finally, a prediction model based on the four factors model is constructed as follows:

(Number of coworking spaces in the block group) = -0.123373 + 0.109078*(Number of Museums) + 0.055948 * (Number of Groceries) + 0.153472 * (Number of Coffee shops) + 0.093279 * (Number of subway station within 0.1mile buffered area of the block group)

Use the above model to predict the number of coworking spaces in the block, and we get a map (Map #9 in Appendix-1) showing the Std. Residuals. The red blocks mean the number of coworking in real life is more than the prediction while the blue blocks mean the number of coworking in real life is less than the prediction by the model. Theoretically, the model could help us to identify the potential blocks to look at for future coworking space development, as the blue blocks currently have much less coworking spaces in the area than the prediction.

Conclusions and Future study

This exploratory study of coworking space revealed a series of correlations between the number of coworking spaces and the urban demographic factors, parks, facilities, eating and drinking places, and public transportations. Statistically significant correlations are found between the location of coworking spaces and the location of the coffee shops, groceries, drinking spaces, fitness centers, museums and subway stations. The result suggests that the current coworking spaces are not randomly distributed, they are significantly clustered and are mostly located in a mixed-use area, so that they could take more advantages of the easy access to the surrounding facilities and urban resources.

This study is helpful for understanding the unban factors that explain how coworking spaces are clustered and why they locate in some locations but not in others, as a thread to interpret coworking as a new urban lifestyle. From a developer's perspective, the location significantly reflected the consumer and market's preference in the characteristics of coworking space's urban environment. For city planners, if coworking space is considered as the incubator for startups and self-employed workers, it could also be regarded as a kind of social infrastructure which provides access to a flexible, affordable and well-connected workplace. The central location in the city could contribute to the livability of the downtown area, as it attracted people to the center of the city and it could provide affordable working environment for workers in different social-economic status.

A weakness of this research reflected among the process is the level of block groups. As the block's size varies, the number of facilities in the block's boundary might not reflect the actual density of the facilities. Moreover, the ACS census data's reliability in block groups level also threatens the validity of this research. Similar research could be done in Census tract level to see if there are similar correlations exist among the variables. How coworking spaces contribute to the mixture of people in different social-economic status, and how it affects the urban environment could be an interesting study topic for the future. Future research could also include the factors such as land-use, density, and other zoning regulation factors, to study the relationships between zoning controls and think about how policies could encourage or obstruct the development of the coworking spaces. A closer look of the consumers, a market analysis, could also help to reflect more meaningful information about the change of demography, workforce, employee's preference and meaning of work, in the coworking space consumer population.

Appendix-1

Map #9: Prediction based on model

References

Atkinson, R. (2004). The evidence on the impact of gentrification: new lessons for the urban renaissance?. European Journal of Housing Policy, 4(1), 107-131.

Barley, S. R., & Kunda, G. (2004). Gurus, hired guns, and warm bodies: Itinerant experts in a knowledge economy. Princeton University Press.

Banerjee, M., Tolbert, P. S., & DiCiccio, T. (2012). Friend or foe? The effects of contingent employees on standard employees' work attitudes. The International Journal of Human Resource Management, 23(11), 2180-2204.

Deskmag. (2018, February 22). 1.7 Million Members Will Work in Coworking Spaces by the End of 2018. Retrieved November 25, 2018, from http://www.deskmag.com/en/1-7-million-members-will-work-in-coworking-spaces-by-the-end-of-2018-survey

Deskmag (2013). ‘The history of coworking in a timeline’. [http://www.deskmag.com/en/the-history-of-coworking-spaces-in-a-timeline]

Deskmag (2017a). ‘2017 global coworking space survey’. [https://www.slideshare.net/carstenfoertsch/utilization-of-coworking-spaces-members-of-coworking-spaces-part-2-of-2-80912960]

Deskmag (2017b). ‘More than one million people will work in Coworking Spaces in 2017’. [http://www.deskmag.com/en/the-complete-2017-coworking-forecast-more-than-one-million-people-work-from-14000-coworking-spaces-s]

Eliot B. (2017 Oct., 19), WeWork: A $20 Billion Startup Fueled by Silicon Valley Pixie Dust, The Wall Street Journal.  Retrived from: https://www.wsj.com/articles/wework-a-20-billion-startup-fueled-by-silicon-valley-pixie-dust-1508424483

Freelancers Union (2012) Freelancing in America. A national survey of the new workforce. New York.

Friedman, G. (2014). Workers without employers: shadow corporations and the rise of the gig economy. Review of Keynesian Economics, 2(2), 171-188.

Gandini, A. (2015). The rise of coworking spaces: A literature review. Ephemera, 15(1), 193–205.

Hall, M. (2018, September 19). Coworking Made Up Nearly One-Third Of Manhattan’s Office Leases In August. Retrieved September 21, 2018, from https://www.bisnow.com/new-york/news/office/coworking-made-up-one-third-of-manhattans-office-leases-in-august-93026

Hot in the City; There's been a marked shift in population trends, with increasing numbers of millennials and baby boomers alike ditching the suburbs for an urban lifestyle. What does it mean for banks? (2014, April 1). American Banker Magazine, 124(03), 18. Retrieved from http://link.galegroup.com.proxy.library.cornell.edu/apps/doc/A363465777/AONE?u=nysl_sc_cornl&sid=AONE&xid=7801aac8

Jeszeck, C. A. (2015). Contingent Workforce: Size, Characteristics, Earnings, and Benefits. US Government Accountability Office.

JLL (2017). ‘Shared workspaces: The market perspective’. [http://www.us.jll.com/united-states/en-us/research/property/office/coworking-space-the-landlord-perspective]

Merkel, J. (2015). Coworking in the city. Ephemera, 15(1), 121–139.

McDonald, N. C. (2015). Are millennials really the “go-nowhere” generation? Journal of the American Planning Association, 81(2), 90-103.

Moos, M. (2016). From gentrification to youthification? The increasing importance of young age in delineating high-density living. Urban Studies, 53(14), 2903-2920.

Moos, M., Pfeiffer, D., & Vinodrai, T. (Eds.). (2017). The Millennial City: Trends, Implications, and Prospects for Urban Planning and Policy. Routledge.

Moriset, B. (2010, April). Developing the digital economy in France's rural regions: A new era for telecenters?. In Annual Meeting, Association of American Geographers.

Moriset, B. (2014) ‘Building new places of the creative economy. The rise of coworking spaces’, proceedings of the 2nd Geography of Innovation, International Conference 2014, Utrecht University, Utrecht (The Netherlands).

Rymarzak, M., & Siemińska, E. (2012). Factors affecting the location of real estate. Journal of Corporate Real Estate, 14(4), 214-225.

Spinuzzi, C. (2012). Working alone together: Coworking as emergent collaborative activity. Journal of Business and Technical Communication, 26(4), 399-441. Darchen, S. (2016). “Clusters” or “communities”? Analysing the spatial agglomeration of video game companies in Australia. Urban Geography, 37(2), 202-222.

The Economist (2016, January 23). The millennial generation Young, gifted and held back – The world’s young are an oppressed minority. Unleash them.

Tolbert, P. S. (1998). Two‐tiered faculty systems and organizational outcomes. New directions for higher education, 1998(104), 71-80.

Union, F., & Upwork (2017), Freelancing in America: 2017.

Union, F., & Desk, E. O. (2012). Freelancing in America: A national survey of the new workforce. New York.

Wang, B., & Loo, B. (2017). Hubs of internet entrepreneurs: The emergence of co-working offices in shanghai, china .Journal of Urban Technology, 24(3), 67-84.

Watters, E. (2003). Urban Tribe: A Generation Redefines Friendship, Family, and Commitment. New York: Bloomsbury.

Zook, M. (2008). The geography of the internet industry: Venture capital, dot-coms, and local knowledge. John Wiley & Sons.