NAFTA has increasingly become politicised in the US in recent years, and
with this, there has been heightened speculation on its purpose for ordinary
Americans and if it is in their best interests after all. Using US Census data,
Gaston and Trefler’s (1997) regression model is then applied to estimate the effects
of NAFTA on the employment and wage growth of 14 US manufacturing sectors in
years 1994-2015. IPUMS is then used to measure these effects by geography at a
county-level.
The results show a clear relationship between NAFTA’s tariff reductions
and employment losses, with highly-tariffed industries pre-NAFTA experiencing
the biggest declines in employment levels and slowest wage growth. The findings also align well with
Heckscher-Ohlin’s (1967) trade theory, showing a clear link between trade
liberalisation and stagnant wage growth in some sectors, with migration from
Mexico magnifying these effects. The rust belt is then identified as the region
which has felt these adverse effects the most, with sunbelt states subsequently
profiteering in a North-South manufacturing shift.
The study then makes recommendations on the re-implementation of trade
tariffs with both Mexico and Canada, finding that this would increase
employment levels and stimulate wage growth.
In 1990, the North American Free Trade Agreement (NAFTA) was declared and agreed in principle by 1992, finally ratified under President Clinton’s administration in 1994, to the dismay of many sceptical Americans, including political contender Ross Perot. NAFTA removed most tariffs between the US, Canada and Mexico and liberalised trade amongst the new bloc. Consequently, many corporations took advantage of the cheaper supply of labour in neighbouring Mexico and here began the relocation of plants and manufacturing facilities, resulting in many job losses. Critics of NAFTA argue that the agreement was introduced only to benefit multi-national corporations, bolstering their outsourcing capabilities and reducing costs; all at the expense of the American working classes. These feelings of resentment have only grown, evident with the nomination of President Trump in 2016. With this, came key announcements regarding NAFTA, the first a new 20% tax on all imported lumber from Canada, a clear deviation from NAFTA principles; and the second, a complete renegotiation of NAFTA with Mexico and Canada. Subsequently, this dissertation will investigate the impact NAFTA has had on manufacturing employment and wages since its induction, also exploring if specific regions in the US have been disproportionately affected by the agreement.
Applied research will be used for the dissertation, utilizing Gaston and
Trefler’s (1997) regression model. This will calculate what effects different
NAFTA independent variables have had on manufacturing employment and wages.
These variables vary from concessionary tariff rates to levels of migration. Before
this, an in-depth review of existing literature will first be explored,
commencing in chapter 2.
Studies investigating the effects of NAFTA have proven drastically different results and conclusions, with the overall consensus of critics regarding the agreement being very divided. Going back to the inception of NAFTA, Nystrom et al. (1995) set about creating a strategic cost analysis tool for companies planning to exploit the new trade deal, whilst Wylie (1995) made estimates for the potential implications that NAFTA could have on North America’s manufacturing trade with the rest of the world. Both critics offer a stark contrast in viewing the agreement, one more positive in the interest of multi-national corporations, and the other vastly direr for blue-collar workers in the manufacturing industry. These opposing sides of the spectrum set the tone for the literature review, which will use work from a wide range of authors on NAFTA and evaluate how these findings contribute to the existing academic research available.
Predictions on the effects of NAFTA on wages and employment in the
manufacturing industry were divisive even in its initial stages, with Hufbauer
and Schott (1993) estimating that in its first year, NAFTA would produce a net growth
of 171,000 jobs, whilst Koechlin and Larudee (1992) projected a 490,000 loss in
US manufacturing jobs in years 1992 to 2000, caused by a shift in investment to
Mexico. (Cited in Sunthonkhan, 2010, p.9) In addition, Weintraub (1992) and
Brown (1992) saw NAFTA as being of little significance to overall US GDP
growth, with surveys suggesting that a maximum of 0.5% growth could be expected
in the US, 2% in Canada and 7% in Mexico. Similarly, Hinojosa-Ojeda et al. (2000) also highlighted the minor
impact of NAFTA on US manufacturing employment, finding that trade with NAFTA
partners provided net 23,220 jobs per year.
Wylie (1995) goes a step beyond this, using Houthakker and Magee’s
(1969) estimates of elasticity of import demand for manufactured goods in the
US and NAFTA-related GDP growth predictions to project total import growth.
Subsequently, Wylie’s (1995) calculations pointed towards a 0.85% growth in
imports, suggesting that NAFTA-induced employment and trade growth outweighs
the negative effects of investments moving, primarily, to Mexico. Whilst Wylie attempted to make growth
estimates more dynamic and therefore reliable by using elasticity of import
demand to measure future trends, his work remained incomplete due to a number
of omitted factors in his calculations. For example, he did not consider how
NAFTA could reduce business environment uncertainty, and how as a result, more
US corporations would be stimulated to undertake trade and production
rationalisation, also possibly resulting in companies increasing their vertical
integration capabilities in new, accessible markets. As a result, intermediate
goods traded within NAFTA markets would see high levels of growth due to
corporations utilizing multinational plants and manufacturing facilities to
achieve further economies of scale. Wylie (1995) opted to use Baldwin and
Murray’s (1977) partial equilibrium economic model for trade creation as well
as Bhagwati’s (1992) and Kreinin and Plummer’s (1992) models for trade
diversion. These measured only static production levels, not taking into
account dynamic production rationalisation. (See
appendices) Consequently, the validity of the results can be questioned as
the evidence cited by Wiley (1995) does not measure the trade effects of NAFTA
accurately.
The complexity of predicting the effects of NAFTA on trade, employment
and wages in the US was not fully comprehended by researchers, and there is still
no clear consensus on the impact of NAFTA on the US, even with 20 years’ worth
of data. This is evident in Kondonassis et al. (2008), which analyses 22 manufacturing
industries of the two-digit Standard Industrial Classification (SIC),
estimating that up to 500,000 jobs have been lost in the US a result of NAFTA,
with a large bulk of these being highly paid manufacturing jobs. Despite this,
they argue that the trade liberalisation has not affected wage levels in the
US, a point which is contended by Polaski (2006), who suggests that employee
bargaining power had been reduced as a result of NAFTA, negatively affecting
their wage growth. Similarly, Lawrence et al. (2003) argues that trade
liberalisation has impacted less educated, blue collar workers in the
manufacturing industry the hardest, who according to estimations, have seen a
12% decline in real wages since 1973. Mclaren and Hakobyan (2015) results
develop on this, finding that blue collar workers in the worst affected
manufacturing industries saw significantly slower wage growth (Welfare gains
estimated at 0.2%) compared to other sectors. They continue by linking income
equality to NAFTA, finding that while manufacturing workers had seen their
wages significantly decrease, US corporations witnessed an 88% in profit growth
in the 1990’s, with CEO wages also increasing by over 460%.
Other researchers observe that shrinking wages in manufacturing are
caused by technological advancements, not trade liberalisation. Lavoie and
Therrien (1999) suggest that low skilled manufacturing employees have been
mostly displaced by high-tech machinery, reducing demand for these workers,
also seeing a shift to high-skilled labourers instead. Conversely, Wood (1995),
Cline (1997) and Leamer (1996) examine a connection between trade
liberalisation and declining wages of manufacturing workers. (Cited in Yasin,
2009, p. 48) These conclusions in general are aligned with the Heckscher-Ohlin
trade theory, which proposes that increased import competitiveness reduces
demand for domestic unskilled manufacturing workers by way of cheaper products
produced by less costly unskilled labour the exporting country. (Ohlin, 1967)
This is reinforced by Swagel and Slaughter (1997), who suggest that NAFTA
increases the supply of unskilled labour in the US, therefore reducing wages
for domestic, low skilled manufacturing workers through outsourcing or through
accelerated immigration of unskilled labourers from Mexico into an increasingly
competitive US labour market.
Whilst many of these authors all agree on the same basic conclusion, the
models and concepts used to reach them are all different, for example Wylie
(1995) uses a partial equilibrium model adapted from Baldwin & Murray
(1977), while Yasin (2009) uses factor-content and price-effect approaches, indicating
that these results are not interchangeable with each other as each model can
give a different outcome even with the exact same data.
According to Holmes (1996), the historical location of the manufacturing
belt in the North occurred simply by chance, and inevitably, manufacturers will
move South to more favourable natural climates for production. He uses the
cotton textile industry to study this effect, finding that the lowest quality
producers begin to re-locate first to exploit better growing conditions and
increased cost savings. Subsequently, other firms follow suit to achieve
similar effects while suppliers also begin to re-locate to meet increased
demand in new Southern territories. Holmes’ (1996) outline of firm re-location
can also be extrapolated for use on NAFTA, as this opened up trade with Mexico,
essentially broadening the Southern border for firms looking to relocate. Figure
1 shows a map of the traditional manufacturing regions.
Holmes’ theory is reinforced by Casetti’s (1984) findings, which show that
employment in manufacturing saw a sharp drop in the rust belt region and New
England in the 1970’s, whilst at the same time, Sunbelt regions in the South
saw a rapid increase in manufacturing employment, signalling a clear re-location
of manufacturing from North to South. This trend accelerated throughout the
decade, with later work from Casetti & Jones (1987) confirming this, also
suggesting that the rustbelt to sunbelt swing actually began in the 1960’s. Logan
(2008) reaffirms this in Figure 2, signalling
a clear shift from manufacturing belt states to sunbelt states.
However, there is evidence of selective data bias in Holmes’ (1996),
this is evident in his decision to analyse the cotton industry for his
research, an industry whose production he would have known would be affected by differing climates. If he analysed a
more diverse set of industries, such as the automobile or air conditioning
industries, his results would have been vastly different, thus not aligning
with his pre-determined narrative. Similarly, Glasmeier & McCluskey (1987) suggest
that reasons for production relocation are industry-specific and that a natural
shift to better climates in the South is not a primary reason for this. Yet,
despite its flaws, Holmes’ (1996) cotton example can also apply to this theory
as well, as this proves that each industry is unique in its requirements; cotton
production needing a favourable climate for better yields. Glasmeier &
McCluskey (1987) also found that automobile manufacturers moved their parts
assembly operations away from the rust belt and closer to vital markets to the
South and West. Glasmeier & Leichenko (1996) develop on this by observing a
correlation between increased trade liberalisation and relocation of
manufacturing from the rust belt to the South, where supply of unskilled labour
was much larger, beginning with Mexico’s entry into the GATT in 1985. (General
Agreement on Trade and Tariffs) As a result, the rustbelt and Midwestern states
suffered the most in manufacturing employment and wages, harbouring most of the
3m job losses in years 1979-1992, while Southern regions accounted for only 3%
of these. Yet, this trend only repeated itself once more and was accelerated by
NAFTA, seeing jobs leave the rust belt as well as Southern regions, beyond US
borders, to Mexico.
Yoskowitz et al. (2002) investigate this theory, examining levels of
manufacturing employment in 7 Southern Texas counties between 1990 to 1997. Their
findings indicate that these seven border counties had experienced much higher
levels of employment growth pre-NAFTA, crediting Mexico’s 1985 entry into the
GATT (General Agreement on Tariffs and Trade) for these previously high levels
of growth. However, despite the sudden drop in employment growth post-NAFTA,
Yoskowitz et al. (2002) do not associate the two with each other, instead
citing the reduced growth as a natural decrease from 1985’s high. They also
suggest that counties do not benefit significantly from their proximity to the
border as most of the trade done here moves through these regions and into
bigger cities, such as Houston.
This is contended by McCallum’s (1995) research, which observed 30 US
states and 10 Canadian regions along the joint border, investigating for
location effects of producers within close proximity to the border. It found
that counties located close to the border benefitted from increased trade with
their neighbours, as did producers, concluding that this can also be applied to
the US-Mexico border, expecting even more impressive trade patterns due to
differences in resources and labour. Hanson (1995) also looks at how corporations
are increasingly moving assembly plants overseas, specifically to Mexico. He
predicted that NAFTA tariff reductions would result in firms moving plants and
factories to regions near the US/Mexico border for transportation cost saving
reasons. He also finds a correlation between offshore assembly firms and
employment growth in border towns, consistent with McCallum’s (1995) research.
An adapted form of Gaston & Trefler’s (1997) regression model for
employment and wages will be used for the first objective’s research, testing
for the manufacturing employment and wage effects of NAFTA tariff reductions
between 1994-2015. The model used is
shown below:
yis = θs + βCA∆τ + βUS∆τ + εis,
The equation shows yis as the
result for industry wages and employment, whilst βCA∆ is for industry-specific independent
variables; βUS∆ demonstrates independent variables which can
be used across all industries.
Tariffs against Mexico and Canada are the most important variables
tested in this research, these will be dependent on each industry. As tariffs
have all but been removed between NAFTA countries, tariff concessions will be
used to estimate the present-day effects they would have on employment and
wages. They will be calculated using Schott’s (2004) method of dividing total
duties amassed by the overall taxable value of imports. Data used for these
calculations will be from the US Census Bureau and the Centre for International
Data. (2006) A negative relationship between removing tariffs and manufacturing
employment is expected, whilst wages in high skilled jobs are anticipated to
grow as a result of a higher supply of low skilled labour.
Trade flows to and from Canada and Mexico will also be variables, these
will use import and export values to the US, quantifying any changes in trade
between NAFTA participants. Naturally, it is expected that removing trade
barriers such as tariffs has resulted in higher levels of trade with Mexico and
Canada after NAFTA. It is likely that Canadian imports have impacted US
manufacturing employment and wages positively, due to the similarities between
both countries capabilities in labour and resources. Imports from Mexico are
thought to show a negative correlation with employment, more so than on wages,
specifically damaging low skilled jobs, with a high number of these outsourced.
Migration variables from Mexico and Canada are used to investigate the
effects of imported labour from both nations, with high skilled labourers
expected from Canada and low skilled from Mexico. Both legal immigrants and
estimates of illegal immigrants are included in the data, with illegals likely
to be composed of low skilled workers. Migration from Mexico is expected to be
positive for employment growth as there would be more supply for low skilled
workers, but for the same reason, Mexican migration is expected to be a
negative factor for wage growth. Canadian migration is projected to have an
indifferent effect on wages and employment.
The next variable is labour productivity; this is used to measure both
wage growth and employment levels. As tariffs and supply of labour are made
more accessible, productivity is expected to increase and with this wages, both
at the expense of employment. Data is extracted from the IPUMS (Integrated
Public Use Micro Data Series) database for this.
As US GDP increases, it is predicted that demand for labour follows the
same direction, increasing wages and employment. Whilst a decrease in GDP would
mean a decline in employment due to less capital expenditure available to
companies and lower wages. Gaston and Trefler (1997) also suggest that interest
rates can affect employment levels, with high rates acting as barriers to
corporations looking to borrow capital to expand their operations.
14 2-digit SIC manufacturing sectors were chosen for the study, with
data collected for years 1994 to 2015, the IPUMS database was used to compile
the data. See Table 1. (IPUMS,2017)
For the research objective regarding geographical factors and how they
affect NAFTA’s impact, the sample data was collected from all 3,007 US counties
for employment, whilst only partial county data was used for analysing wages,
with the remaining data used from state statistics from the US Census Bureau.
County shares of manufacturing earnings and employment were collected from the
US department of commerce online resources. (USDC, 2017) The research is
expected to align with previous literature, showing a concentration of
employment losses in the traditional rust belt region of the US.
Most notably, results in Table 1 suggest that tariffs with Mexico
could impact employment positively, whilst small changes in tariffs with Canada
would affect imports but not affect employment significantly. The results
indicate that exports to Canada supported employment slightly, whilst an
increase in imports did the opposite in a big way, negatively impacting
employment considerably. These effects were perpendicular to those for Mexico,
whose imports were the most important growth factor for manufacturing
employment, while exports to Mexico were damaging employment rates. Also, as
anticipated, Mexican migration to the US had an adverse effect on manufacturing
employment, whilst Canadian migration continued to stimulate employment.
Table 2 measures NAFTA effects on manufacturing wages. The results
show that exports to Mexico had the most weight on influencing higher wages,
also indicating a positive relationship between the two. However, imports from
Mexico signalled a more negative impact, though the findings also suggest that
Mexican migration increased average manufacturing wages, contrary to much
research. Conversely, Canadian migration is shown to affects wages negatively, also
finding that tariffs with Canada would reduce wages. The evidence also shows
that manufacturing wages were especially sensitive to changes in tariffs with
Canada, with results indicating that a hypothetical 1% reduction in Canadian
tariffs could increase wages by 3.3%.
Subsequently,to build on these findings, Table 3 looks at how different levels of tariffs affected
employment in manufacturing industries after NAFTA.
Source: Dataset from IPUMS
The findings from Table 3 show employment figures from 1994-2015 for 14 two digit SIC
industries. For low tariff manufacturing industries, there was an average job
loss of 129,142, equal to a 15% decrease in employment after NAFTA came into
effect. Moreover, NAFTA tariff cuts were made proportionally, meaning that the
most highly tariffed sectors would be impacted the most. As a result,
previously highly tariffed sectors lost on average 19.3% more job losses than
low tariff sectors, seeing a 34.3% average fall in employment since 1994, with
the clothing and textiles industries suffering 82% and 72% reductions, in the
same order. It is also notable that these two sectors are composed of mostly
low-skilled labour, meaning that the jobs lost were most likely outsourced to
Mexico, where corporations can cut costs per employee and benefit from NAFTA
tariff cuts. Only the food production and metals fabrication sectors were able
to withstand NAFTA’s tariff adjustments, with an average of 0.8% employment
growth between the two.
Source: Dataset
from IPUMS
In
line with poor employment performance, the clothing production sector also
experienced the lowest wage growth in all 14 analysed industries, increasing
only $627 or 4.1% within a 21-year timeframe. This suggests that as well as
outsourcing clothing production, companies have also seen cheaper, imported
low-skilled labour take up positions within the US, significantly impacting
wages negatively. In comparison, other sectors saw average wages grow by $11,294,
with low tariff sectors seeing wages grow by 8 percentage points more than that
for high tariff sectors. In addition, the data in Table 4 shows that the NAFTA
tariff cuts significantly supported the Lumber industry’s wage growth at 69%.
Canada is the primary exporter of Lumber to the US, suggesting that tariff cuts
between the two has fuelled wage growth beyond that of all other sectors, hence
a positive correlation between the two, as seen also in Table 2. Furthermore, low-tariff industries show the strongest
levels of correlation between job losses and wage growth, whilst high tariff
sectors show more mixed results. Also, the metals fabrication and food
production industries have been the most resilient since NAFTA’s inception,
showing stable signs of employment and high levels of wage growth.
Figure 3 is a map of 3,007 US counties, measuring rates of
employment between NAFTA’s commencement in 1994 to 2015, using data from the US
Census Bureau and IPUMS. As expected, the biggest job losses were experienced
in the manufacturing belt, an area stretching from Wisconsin to Western New
York, whilst the greatest gains were seen in the Southern states of Texas and
Florida.
Counties most affected by
NAFTA were located in Wisconsin, Michigan, Indiana, Pennsylvania, Ohio and New
York, which saw an average manufacturing employment reduction of 29%, or 89,437
jobs. The biggest decline in manufacturing employment was seen in Noble county,
Indiana, with a 59% drop in employment. Brewster county, Texas and Marion
county in Florida saw the highest levels of manufacturing employment growth,
averaging at 17.3% between them, with generally higher levels of growth
correlated with the Southern geographic areas. Another reason for this is that
the rust belt states were very reliant on manufacturing for employment, meaning
that they would be affected the most by tariff adjustments. Figure 4 conveys share of manufacturing
employment per county. There is a relatively high correlation between a
county’s share of employment from manufacturing and the ensuing job losses
experienced in years 1994-2015, with prime examples Indiana and Ohio, who had
27 and 19 counties each with 20% or more jobs in manufacturing. Of these
manufacturing-reliant counties, more than 80% of them were rural.
Southern states, on the
other hand, experienced an influx of corporations moving in with plants and
factories to exploit cheaper low skilled labour and to take advantage of
proximity effects near new supply chains, thereby bolstering the effects of
vertical integration.
Figure 5’s results are
concurrent with those in Table 4, with 26 states seeing manufacturing wages
grow by 41-45%, 13 states 35-40%, 10 states 46-55% and Texas with a 57.1%
average wage growth since NAFTA came into effect. In line with the other
results, wage growth is generally greater in regions with most job losses, more
specifically the rust belt region, with Ohio, Michigan, Pennsylvania, Indiana
and Wisconsin collectively seeing an average manufacturing wage growth of
43.5%. Moreover, the Southern states of Texas and Florida once again see
positive signs of growth, with a 53.3% averaged wage growth between them.
Figure 5 Manufacturing Wage growth by state; 1994-2015
Wage growth shows some
signs of correlation with the average share of wages coming from manufacturing
in each county. 48% (1465) of all US counties had earnings from manufacturing
account for 10% or more of their total incomes, with a majority of these in the
East and South Eastern pockets of the country and Texas in the South. See Figure 6
NAFTA’s impact on US manufacturing employment and wages
The findings of this
research regarding the trade agreement’s impact on manufacturing employment and
wages suggest that the trade liberalisation pacts of NAFTA have affected
previously high-tariffed industries the most, with these sectors experiencing
the lowest levels of wage growth and shedding the highest number of jobs in
years 1994-2015. The regression results indicate that re-introducing tariffs with
Mexico could slow employment losses down and increase wages in the most
severely affected sectors, such as clothing production and textiles
manufacturing. Employment and wages in low-skilled labour abundant
manufacturing sectors have also been harmed by Mexican migration and
outsourcing, which has allowed corporations to exploit cheaper labour. This is consistent
with Heckscher-Ohlin’s trade theory. (Ohlin,1967) Unexpectedly, the results
also show that imposing tariffs against Canada would increase employment
levels, especially in highly traded US-Canada sectors such as Lumber. This
would also soften the negative effects of Canadian imports on US employment
levels, however, wages would suffer as a result of tariffs with Canada.
Overall, the conclusions align with those from Swagel and Slaughter (1997),
Cline (1997) and Leamer (1996), whose findings also bring into question the
benefits of NAFTA for the United States.
Regional effects of NAFTA
The results confirm prior
predictions that the rust belt region was hit the hardest by NAFTA,
experiencing the highest job losses in the US. It also found that counties with
the highest shares of earnings and employment attributable to manufacturing
were worst affected. Many of these were near to the Canadian border, dispelling
Yoskowitz et al’s (2002) theory of favourable location effects of counties
closer to the borders of trading partners. However, the research finds that the
rust belt region recovered well in regards to wage growth, seeing an average
43.5% wage growth since 1994. Moreover, the results also show a clear shift in
employment from rustbelt to sunbelt states, concurrent with Logan’s (2008)
illustrations and Glasmeier and McCluskey’s (1987) work, which suggests that
manufacturing has been moving South as a result of trade liberalisation and a
larger supply of cheap, unskilled labour. Subsequently, the Sunbelt states of
Texas and Florida were found to have benefitted the most from trade
liberalisation, seeing the highest levels of manufacturing employment and wage
growth, possibly due to the influx of cheaper imported labour from Mexico in
the 1990’s.
The main limitations
experienced in the study lie with difficulties in data collection and time
restraints which meant that further exploratory research could not be
conducted. To develop, whilst county-level data was collected, compiled and
mapped out for employment, wage growth data on a county-by-county basis was
unattainable and only available to US students/scholars. As a result, the wage
growth map was limited to a state level analysis, meaning that results for this
section lacked the specificity of the employment section and consequently, may
be defined as incomplete or even unreliable. In addition, time restraints also
meant that regions could not be analysed to find out which industries had
primarily been causing employment losses. Moreover, recent literature on the
location/geographical effects of trade liberalisation also proved to be
insufficient, with dated work the only viable option.
Additional research is
required to fully comprehend the effects of NAFTA on manufacturing employment
and wages and the regions most affected by this. A recommendation for future
researchers would be to analyse a longer timeframe, perhaps from 1980(Before
GATT and NAFTA). This would allow for a more comprehensive look into the
conditions before NAFTA and give researchers the opportunity to compare before
and after NAFTA, rather than estimating how tariffs would affect employment and
wages in the present day. Also, future research should also explore which
specific industries affected levels of employment within different geographical
regions.
Trade Creation equation – TCi = MP,LSi( T* –
TO) / ( 1 +K”)
Trade diversion equation – TDi = MNPi(TCi/Vi)
You have to be 100% sure of the quality of your product to give a money-back guarantee. This describes us perfectly. Make sure that this guarantee is totally transparent.
Read moreEach paper is composed from scratch, according to your instructions. It is then checked by our plagiarism-detection software. There is no gap where plagiarism could squeeze in.
Read moreThanks to our free revisions, there is no way for you to be unsatisfied. We will work on your paper until you are completely happy with the result.
Read moreYour email is safe, as we store it according to international data protection rules. Your bank details are secure, as we use only reliable payment systems.
Read moreBy sending us your money, you buy the service we provide. Check out our terms and conditions if you prefer business talks to be laid out in official language.
Read more