Two journalists, Dawn Gilbertson and Jonathan J. For some it has been an opportunity to grow and for others a challenge to be met. Barreto, Hector V. Gilbertson, Dawn, and Jonathan J. Hagenbaugh, Barbara. Manufacturing Jobs Fading Away Fast. Jette, Julie. Rowe, Claudia. Department of Commerce.
Bureau of the Census, Foreign Trade Statistics. Retrieved on 17 April Federal Reserve Bank of Dallas. Canas, Jesus, and Roberto Coronado. Retrieved on 18 April Tariff elimination for qualifying products. Before NAFTA, tariffs of 30 percent or higher on export goods to Mexico were common, as were long delays caused by paperwork. Additionally, Mexican tariffs on U. NAFTA addressed this imbalance by phasing out tariffs over 15 years. Approximately 50 percent of the tariffs were abolished immediately when the agreement took effect, and the remaining tariffs were targeted for gradual elimination.
Elimination of nontariff barriers by This includes opening the border and interior of Mexico to U. Nontariff barriers were the biggest obstacle to conducting business in Mexico that small exporters faced. Establishment of standards. The three NAFTA countries agreed to toughen health, safety, and industrial standards to the highest existing standards among the three countries which were always U.
Also, national standards could no longer be used as a barrier to free trade. The speed of export-product inspections and certifications was also improved.
Supplemental agreements. 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.
The Competitive Advantage of Nations
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.
Economic globalization in the global post-crisis of limits and deadlocks
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.