In the specifications that include both individual and country controls, only 9—10 countries out of 56 per globalization dimension had any variation in the globalization quartiles, resulting in a big drop-out of countries from the analysis including some very large ones, e. As this reduction in the sample is attributable to our transformation of the globalization indices into dummy variables which we adopted for ease of interpretability of the resulting coefficients , this can be remedied by avoiding the transformation of the globalization indicators and using them as un-transformed variables.
Hence, when it comes to the important comparison between the OLS- and FE-based results, we will revert to the use of the untransformed variables. As a first step, we deal with the confounding problem by including a set of country-level covariates contained in vector C 2 ct , as in specification 2 below.
The choice of the country-level confounders was informed by the existing literature on the factors which facilitate movement of trade and investment between countries, and therefore are drivers of globalization. In addition, these variables are expected to be related to overweight risk.
The size of a country's GDP is also likely to be related to the level of economic development, and thus, in turn, may affect the obesity risk Goryakin and Suhrcke, In addition, foreign investors may consider it more worthwhile to invest in countries with higher overall levels of education and socioeconomic development Asiedu, ; Walsh and Yu, The Human Development Index HDI developed by UNDP is a well-known metric which takes into account not only living standards as measured by GDP per capita, but also two other important components: life expectancy at birth and the literacy rate.
Globerman and Shapiro , for example, found that HDI and FDI were significantly correlated in specifications which did not control for governance institutions and infrastructure indicators. Likewise, it was found in several studies e. In addition, another important determinant of globalization and potentially of economic and social development, which in turn may affect overweight prevalence independently of globalization is the quality of economic and legal institutions Asiedu, ; Obwona, ; Walsh and Yu, It takes into account a number of factors potentially important in the decision-making by foreign investors to engage in economic relationships with countries, such as: a quantitative measure of the ability to start, operate, and close a business; absence of tariff and non-tariff barriers; measure of the tax burden imposed by government; security of property rights; freedom from corruption; flexibility of the labour markets.
Therefore this indicator is likely to be particularly valuable in our search for relevant proxies for drivers of country-level globalization. As we mentioned above, although the above approach is designed to control for a range of potentially important confounders, not taking advantage of the within-country variation, when such option is in principle available, would be too costly.
Therefore, as a final check, we also conduct country fixed effects estimations on the untransformed globalization scores. Although parameter interpretation is more difficult in this case, there is much more within-variation when untransformed scores are used, and this allows us to test whether findings from the OLS estimation will also hold when controlling for potential time invariant, unobserved country-level confounding.
The authors of the study did not have to obtain ethical approval, as they only analysed secondary, fully anonymized individual-level data from the publicly available Demographic and Health Surveys, as well as some country-level data. In most countries where there were at least two years worth of observations, overweight prevalence tended to increase over the years, although at different rates.
Overweight prevalence was generally considerably higher in Eastern Mediterranean countries, and was the lowest in Africa and South East Asia. In almost all countries, the value of the score increased, although again, the rate of change did differ. It is evident that the most globalized countries e. Turkey, Brazil, Egypt, Jordan tended to remain the most globalized in most years, while the same consistency was true for the least globalized countries e.
There appeared to be more variation in relative ranking for countries that were in between these two extremes, although in most cases the rate of change in the score was modest. Finally, Figs. These figures reveal that the relationship appears positive, quite pronounced and mostly linear for the social globalization score. On the other hand, it appears considerably weaker for the economic score. For total and political scores, the relationship seems quite strong, but mostly non-linear. In the former case, it seems that the association is flat for the least globalized countries, before becoming strongly positive.
For the political dimension, it appears that there is no relationship to overweight for the majority of countries, except for the most globalized ones, for which we observe a strongly positive association. Lowess, unconditional association between overweight and total globalization index, — Lowess, unconditional association between overweight and economic globalization index, — Lowess, unconditional association between overweight and social globalization index, — Lowess, unconditional association between overweight and political globalization index, — In the first column, not controlling for any covariates except for time dummies and the Sub-Saharan Africa dummy, we find that living in the countries which are in the top quartile for this metric is related to a There is also a visible gradient: each higher total globalization quartile is associated with a greater overweight risk, with the shape suggesting a convex pattern.
However, as this association may in part be driven by country-level confounding, it is also important to consider its robustness by including relevant controls. In column 2, the adding of individual control variables improves the precision of the estimates, while also somewhat reducing the magnitude of the association. What matters more, however, is the addition of the country level controls: results in column 3 demonstrate that their addition further reduces the magnitude of the association, although the parameters for the globalization dummies remain significant and positive.
The relationships between the index of total globalization and overweight in women aged 15—49, Ordinary least squares OLS regression results.
Cluster-robust standard errors in parentheses. Sample restricted to women aged 15— No controls except time dummies and Saharan African dummy are included in the baseline specification. Reference categories for each of the sets of dummy variables: living in the least globalized quartile of countries, women with higher education, aged 35—49, having 6 or more children, being unemployed, and living in a rural location. All specifications contain time dummies. Women with no children are less likely to be overweight than women with 6 or more children, whereas women with 1—5 children were more overweight than those with 6 or more children.
Moreover, an increase in the size of the market i. With HDI ranging from 0 to 1, an increase by 0. Interestingly, better economic and legal institutions have an opposite effect: an increase of the score by 1 is related to an about 0. Prior to entering into the regression results, we determined whether each of the sub-components of globalization indeed captured distinct phenomena.
The results in column 1 without controls for any factors except time dummies and a sub-Saharan African dummy, indicate that greater economic globalization is associated with a greater risk of being overweight. Adjusting for individual covariates, however, reduces the magnitude of the association. The biggest impact on parameter sign, however, occurs after adding country controls: now the relationship becomes concave, with people living in the most economically globalized countries having lower probability of being overweight, although this finding needs to be seen in the light of the very small magnitude of this association i.
The relationship between economic, political and social globalization and overweight in women aged 15—49 years, OLS regression results. No controls except time dummies and sub Saharan African dummy are included in the baseline specification columns 1, 4, and 7. In columns 2, 5 and 8, controls also include education, age, number of children, occupation and urban residence dummies. In the basic specification, column 4 , the relationship appears convex, with a fall in the probability of being overweight in the second and third quartile, before an increase for the most politically globalized countries column 4.
However, the addition of individual, and especially country level controls, leads to a more pronounced association: column 6 shows that people living in the most politically globalized countries have a This is also true for people living in the third quartile, although the increase in the probability of overweight is considerably smaller.
It appears that this dimension has the most stable and pronounced association with overweight across dimensions, as adding different sets of control variables changes the magnitude of the association only slightly. People living in the most socially globalized quartile have an about 18 p.
Next, we consider the association between overweight and all globalization indices taken together. On the other hand, putting these scores together in the same model may help ensure an additional degree of control for residual confounding. Some of our findings may be partly driven by the differences in sample size across specifications. Earlier in the paper, the analysis with OLS using globalization scores transformed into quartiles was presented as this allowed a more intuitive interpretation of results. However, we recognize that this approach is costly, as it effectively precludes a country fixed effects analysis which would allow controlling for an important source of unobserved confounding due to a very small within-variation.
Even though interpretation of our key parameter estimates now becomes less clear, this comparison is useful in that it allows us to examine whether the OLS findings continue to hold when the assumption of no correlation between globalization scores and time-invariant unobservables is relaxed. We also see that the magnitude of the CFE associations remains substantive.
For example, a 50 percentage point p. This compares with an about Robustness checks: estimating the relationship between overweight and globalization using original globalization scores. The following controls are added in all specifications: age, number of children, occupation and urban residence dummies, Saharan African dummy, total GDP constant dollars ; Human Development Index, Economic Freedom score.
We also estimate overweight as a quadratic polynomial function of globalization dimensions results not shown here, but available on request. In order to ensure better interpretability and to mitigate the multicollinearity problem, we centred our estimation on the mean values of the globalization dimension scores. We found the main parameters to be virtually identical for all dimensions. In addition, there appears to be a convex relationship between total and political globalization and overweight, a mostly linear negative relationship between economic globalization and overweight, and a mostly linear positive association between social globalization and overweight.
While most of the existing literature focussed on the relationship between economic globalization and obesity, specific quantitative measures of the range of potentially very different globalization-related drivers involved have not been examined previously. In this analysis we find that the relationship between overweight and globalization depends on the specific dimension of globalization. Thus, while both political and especially social globalization dimensions appear strongly positively related to the greater overweight risk, the same is not apparent for economic globalization.
More concretely, comparing different dimensions of globalization and including suitable adjustments for confounders and covariates we find for the first time that political and social globalization consistently show a positive association with the individual odds of overweight: in our preferred specification i. This finding is also confirmed in the models using the untransformed globalization scores, although the magnitude of the association is notably smaller for the social but not for the political dimension in the CFE compared to the OLS model. Although arguably the biggest attention has so far been directed at the impact of economic globalization, we have found that living in the most economically globalized quartile of countries predicts a 1 p.
This is a rather surprising finding, given the focus of most of the literature on the potential link between obesity and economic globalization Hawkes, , and the scant attention paid to other dimensions. Having said that, the parameter sign for the economic dimension was quite sensitive to the inclusion of country-level controls.
This appears to be consistent with the hypothesis that at least part of the relationship between economic globalization and overweight may be driven by country-specific factors such as economic development and infrastructure, education, attractiveness of economies to investors, as well as the size of the market. Inevitably, our study suffers from several limitations.
For example, the sample was necessarily restricted to women only, and mostly of child-bearing age. Nevertheless, since the age group of 15—49 represents the most productive group of women, who also typically have a number of dependants, focussing attention on this demographic segment may be warranted for informing policies to tackle overweight.
Most importantly, we are limited in drawing major causal claims about our findings, especially in relation to 19 countries that were only present in the sample for one year and thus could not provide any within-variation for the fixed effects analysis. There are a few other intrinsic data-related concerns which call for caution when interpreting the findings. This is potentially problematic in that such women may more likely be overweight, although the reverse may be true in the lowest income countries, where both pregnancy and breastfeeding may lead to large energy needs relative to family resources and thus potentially to malnourishment.
Another problem is that very few countries stayed in the sample for all periods, given the nature of the DHS data collection. Whereas in some countries e. Egypt, Ghana data was collected every five years or even more frequently, in many others it was collected for no more than two years.
In 19 countries, data was only available for one year. There was also very little variation in our categorical globalization variable across years, which prevented us from undertaking country fixed effects analysis using the globalization indicator dummies. It remains possible, however, that some time-varying variables which country fixed effects cannot control for may still be a source of bias for our results.
For example, availability of infrastructure, wars, economic shocks and famine may affect both the extent of globalization and overweight risk. However, although we are not controlling for these factors explicitly, we nevertheless control for the Human Development Index, as well as the Index of Economic Freedom which proxies for the quality of economic and legal institutions. Both of these variables, in our view, should to a large extent account for such confounders.
Changing geopolitics and American investment support provided opportunities for many commercially minded cities to become highly specialised. Often benefiting from sound city leadership, cities in this wave of globalisation were successful in restoration and in planning ahead to avoid the worst in congestion. Munich, Toronto and Tokyo are all prominent examples of cities in this wave. These cities were able to achieve intensive clustering and organisational modernisation in a supportive political environment among higher tiers of government.
For this wave, it was key for corporate proficiency to partner with a strong knowledge platform in order to develop successful products for export. Cities in this wave of globalisation were prominent well into the mids, by which time they had acquired many of the core assets that make them competitive today. Although these cities have experienced slowdowns and setbacks in the last 25 years, they have successfully preserved a culture of knowledge and innovation to support specialisations with global reach.
Perhaps the most remarkable example during this wave was Singapore, the only global city also to be a fully autonomous city-state. But its first prime minister, Lee Kuan Yew, was not deterred by geographic limitations and immediately pursued a labour-intensive industrialisation policy that was open to overseas capital. Emphasis was placed on tax incentives for industrial investors, improved labour discipline, technical education and foreign expertise, in order to specialise in industrial technologies. Singapore immediately tried to build trade and connectivity roles by leveraging its position as a multilingual city, with links to east and west, and a strong western legal model.
Since achieving global status, Singapore has continued to adapt to the global marketplace through state-led policies and programmes that encourage light manufacturing and high-tech research investment. Its economic development board and housing development board, in particular, have been able to leverage the scale of large public institutions to create effective systems at a citywide scale. The postwar wave of globalisation was also striking for the process of decolonisation, which saw many cities lose their imperial roles. In the past, many global cities fell along with the empires that spawned them.
Indeed, for London, this period coincided with a challenging postwar process of deindustrialisation, port decline and outdated regulation in its stock exchange, banking and insurance sectors. Although London attracted immigrants from the Commonwealth diaspora, the city continued to lose population until the mids, when it resumed the journey toward becoming a quintessential global city. The decade after the oil crisis and the subsequent global slowdown was a period of great flux in the global economy.
Older models became discredited and new ideas and solutions came to the fore. By the mids, a new wave of globalisation was under way, led initially by a small group of cities but with an increasing number of smaller cities also becoming globally engaged and globally oriented for the first time. For a small, elite group of financial centres, the mids started a wave of resurgence when these cities began to re-attract people, business and capital.
For Asian members of this group, such as Tokyo and Singapore, this wave was characterised by a more selective and tactical globalisation. It promoted internationalisation of financial and business sectors through liberalisation reforms, while also protecting against over-exposure to western cultural influences. In this wave of globalisation, cities also re-established themselves as information and media capitals in their respective regions. At the same time, the new wave of globalisation was showcasing a set of cities that until then had not been globally oriented.
The global geopolitical context had changed considerably as a result of the fall of communism in Europe, the unification of Germany, the opening up of China beginning in , and the Oslo peace accords , part of the Arab-Israeli peace process, in The agreements of new trade arrangements — such as the European Economic Area and the North American Free Trade Association, and the establishment of the World Trade Organisation in — ushered in an era of multilateralism that allowed many countries and cities to globalise.
New tradable specialisations associated with the information and communications technology revolution precipitated the unexpected globalisation of many cities. Bangalore is one of the most striking examples: its electronics specialisation took off in after the arrival of Texas Instruments. This prompted other multinational firms to relocate and boosted the local software design environment, laying the platform for successive climbs up the value chain.
With its tech cluster benefiting from a supportive early stage investor arrangement and a positive attitude to risk, Tel Aviv is noted for its numerous tech inventions. This wave of globalisation was distinctive for a more intentional approach by city governments and the rise of strategic planning for globalising cities.
For cities like Barcelona and Tel Aviv, city governments designed strategic approaches to infrastructure, quality of life, and architecture in order to build a profile that would attract global talent. Many also converted their commercial land into use by more innovation-led activity. In nearly all cases, these cities witnessed an unprecedented wave of population diversification, fuelled by immigration and mobile talent.
By the climax of this wave in , these cities had a totally different global proposition — and new problems that had appeared alongside growth. The most recent cycle of globalisation has seen the start of a new wave of distinctive, specialised and globally aspirational cities. Many of them are higher-income cities within their respective regions and seek to leverage their efficient infrastructure, improved quality of life, and better security and environmental performance compared to the larger megacities.
Cities in this wave are less likely to have major political or institutional functions. Rather, they are competing in more dynamic and fast-moving global industries where opportunities have arisen for greater market share. In the aftermath of the global financial crisis and its impact on public sector finances, the ability of cities to find new sources and tools for investment has been a key enabler for globalising cities since Brisbane is an example of a city that has been able to globalise with the help of a large and financially astute city government, which has used the surplus from its recent commodities boom to erect a more international model of economic development.
The Brisbane city council has pursued a wide range of joint ventures, sponsored business conventions and sporting events, and convinced its state government to prioritise Brisbane for road and rail infrastructure funding. It has also experimented with public-private partnerships and toll roads, with some success. The ability to convey opportunity and appeal to the world, whether to talent, investors, or visitors, has also been very important.
London is an example of a city that has been successful in this area since In medicine, media, and digital industries, the ingredients of leading sector innovations, combined with an open city with deep labour markets and cosmopolitan liveability, have created a winning formula. A comparison of the waves of globalisation in the last two centuries with the earlier waves shows clearly that the duration of each wave is becoming shorter. Where waves once lasted a century or more, they now appear to run their course in as little as to years, and in the future this duration may be even shorter.
As the global economy becomes ever more integrated, globalising city waves increasingly come to resemble global economic cycles, and the windows of opportunity for cities to participate close quickly. Although there are vast differences between the networks of cities along the ancient Silk Roads and the 21st-century system of global value chains and competitive advantage, there are also striking parallels.
Many started out from an unpromising or uncompetitive situation because of either internal weaknesses or external disadvantages. Sometimes cities have endured long periods of global isolation, and begin to internationalise only when geopolitical changes occur and foreign investment arrives.
This was certainly the case in the 20th century, and in the 21st century it is visible in many other cities outside the established west. History shows this is a risk if cities lose competitiveness in traded sectors, fail to embrace innovation or to project influence, are closed to immigration and entrepreneurship, or are unable to adapt to a changing geopolitical or geoeconomic centre of gravity.
Over time, these alternative pathways result in very different kinds of global cities. Fourthly, a focus on networked cross-border dynamics among global cities also allows us to capture more readily the growing intensity of such transactions in other domains—political, cultural, social, and criminal. Global cities around the world are the terrain where a multiplicity of globalization processes assume concrete, localized forms.
These localized forms are, in good part, what globalization is about. Recovering place means recovering the multiplicity of presences in this landscape. It is one of the nexi where the formation of new claims, by both the powerful and the disadvantaged, materializes and assumes concrete forms. Abu-Lughod, J. Minneapolis: University of Minnesota Press, Amen, Mark M.
Archer, and M. Martin Bosman, eds. Bartlett, A. Sassen, ed. New York and London: Routledge, Berner, E. Bonilla, F. Melendez, R. Morales, and M. Borderless Borders: U. Latinos, Latin Americans and the Paradox of Independence. Philadelphia, Temple University Press, Brenner, N. Theodore, eds. Malden, MA: Blackwell Publishers, Bryson, J. Daniels, eds. The Service Industries Handbook. Cheltenham, UK: Edward Elgar, Castells, M. Oxford: Basil Blackwell, Chen, X. In Environment and Planning, 10 : — Ciccolella, P. Mignaqui, I. Copjek, J. Sorkin, eds. Giving Ground: The Politics of Propinquity.
New York: Verso, Farrer, G. Friedmann, J. Giesecke, G. London, Routledge, Gugler, J. World Cities beyond the West. Globalization, Development and Inequality.
Cambridge: Cambridge University Press, Hagedorn, J. Harvey, R. Herzog, L. Kazepov, Y. London: Blackwell, Katznelson, I. Marxism in the City. King, A. Re-presenting the City. Ethnicity, Capital and Culture in the 21st Century.
London, UK: Macmillan, King, D. Kloosterman, R. Rath, eds. Oxford: Berg Publishers, Landrieu, J. May, T. Spector, and P. Veltz, P. Krause, L. Petro, eds. Lloyd, R.
The Impact of Globalization on the Evolution of Urban Systems (Geopolitics and History) - Kindle edition by ELENA MANOLACHE. Download it once and read it. The Impact of Globalization on the Evolution of Urban Systems (Geopolitics and History) eBook: ELENA MANOLACHE: lirodisa.tk: Kindle Store.
Lo, F. Yeung, eds. Emerging World Cities in Pacific Asia. Tokyo: United Nations University Press, Marcotullio, P. Marcuse, Peter. New York: Oxford University Press, Parnreiter, C. Peterson, M. Ribas-Mateos, N. Somerset, NJ: Transaction, Rutherford, J. Aldershot: Ashgate, Sandercock, L. Sassen, S. Smith, ed. Beverly Hills, CA: Sage, , — New upd.
Sennett, R. New York: W. Norton, Smith, R. Taylor, B. Derudder, P. Saey, and F. Witlox, eds. London: Routledge, , — Stren, R. Cohen, B. Ruble, J. Tulchin, and A. Garland, eds. Washington D. Taylor, P. Taylor, and P. Knox, eds. Cambridge: Cambridge University Press, , 48— Witlox, F. Cities in Globalization: Practices, Policies and Theories. London: Routledge, Torres, R. Miron, and J. Inda, eds. Race, Identity and Citizenship. Malden, MA: Blackwell, On the other hand, for a variety of less system-centered view of cities see, e.
See also Stren A central proposition here, developed at length in my work, is that we cannot take the existence of a global economic system as a given, but rather need to examine the particular ways in which the conditions for economic globalization are produced. This requires examining not only communication capacities and the power of multinationals, but also the infrastructure of facilities and work processes necessary for the implementation of global economic systems, including the production of those inputs that constitute the capability for global control and the infrastructure of jobs involved in this production.
The emphasis shifts to the practice of global control: the work of producing and reproducing the organization and management of a global production system and a global marketplace for finance, both under conditions of economic concentration. The recovery of place and production also implies that global processes can be studied in great empirical detail. We are seeing the formation of an economic complex with a valorization dynamic that has properties clearly distinguishing it from other economic complexes whose valorization dynamic is far more articulated with the public economic functions of the state, the quintessential example being Fordist manufacturing.
This in turn brings up a question of control linked to the currently inadequate capacities to govern transactions in electronic space. In this sense, global cities are different from the old capitals of erstwhile empires, in that they are a function of cross-border networks rather than simply the most powerful city of an empire.
There is, in my conceptualization, no such entity as a single global city as there could be a single capital of an empire; the category global city only makes sense as a component of a global network of strategic sites. The corporate subsector which contains the global control and command functions is partly embedded in this network.
There are multiple specifications to this argument. For instance, and going in the opposite direction, the development of financial instruments that represent fixed real estate repositions the latter in various systems of circulation, including global ones. In so doing the meaning of capital fixity is partly transformed and the fixed capital also becomes a site for circulation. For a fuller elaboration see Sassen , ch. There is a fourth case which I have addressed elsewhere Sassen , ch. A central urban hierarchy connects major cities, many of which in turn play central roles in the wider global system of cities: Paris, London, Frankfurt, Amsterdam, Zurich.
And then there are several geographies of marginality: the east-west divide and the north-south divide across Europe as well as newer divisions. In Eastern Europe, certain cities and regions, notably Budapest, are rather attractive for purposes of investment, both European and non-European, while others will increasingly fall behind, notably in Rumania, Yugoslavia, and Albania. We see a similar differentiation in the south of Europe: Madrid, Barcelona and Milan are gaining in the new European hierarchy; Naples, Rome, and Marseille are not.
For a general overview of European cities see Kazepov This also holds in the highly developed world.