Defining the trend: Cement consumption vs GDP

Growth in Gross Domestic Product (GDP) per capita, a measurement of the average national standard of living, can be a contributing factor to cement demand. Increased industrialisation caused by economic expansion has a tendency to drive corresponding increases in cement consumption. This relationship is well known and has been widely used in the past to both judge the relative economic growth between nations and forecast likely cement consumption rates as a given nation's GDP increases. It is this latter use that the current article focuses on, examining the relationship of the two variables in detail across multiple years to assess the relevance of the measures as an accurate forecasting tool.

The GDP versus Cement Consumption graph (See Figure 1) is a striking visual representation of a country's stage of development. Emerging nations, such as China, South Korea and Saudi Arabia, are instantly recognisable. Their cement demand is often far in excess of countries with comparable GDPs and is representative of the significant national investments in infrastructure taking place in these areas.

By contrast, smaller economies with low GDPs and little or no infrastructure investment have small cement consumptions and densely populate the lower end of the graph. The more established economies, for example the US and those in Western Europe, have cement consumptions in line with these less affluent nations since large investments in infrastructure and urbanisation have already taken place.

The relationship between GDP and cement consumption, plotted by multiple sources and years, shows a general inclination towards a cement consumption of 600kg per capita or less in nations with per capita GDPs in excess of US$25,000. This pattern is usually represented by a trend line with a steady incline that reaches a plateau or declines gradually once the GDP reaches that threshold. In most years, the majority of data points can be fitted to this model, however, there are a few regular exceptions.

Above - Table 1: GDP/capita (US$) and cement consumption (kg/capita) for given years.

Exceptions to the rule in 2012

Singapore: Singapore has been consuming cement at very high rates for a considerable period of time. With cement output so low it is negligible, cement represents a significant proportion of national imports.¹⁰

Cement consumption in the country reached 958kg/capita in 2002, over three times the world average for that year, although the GDP had yet to breach the US$25,000 mark. When the economy finally exceeded US$25,000/capita in 2005, cement consumption had decreased considerably and barely exceeded the observed 'limit' of 600kg at 690kg/capita. The same was true in 2007, although growth in multiple sectors¹⁸ meant that GDP continued to increase significantly to US$36,766/capita.

2008 was the first year in which Singapore truly placed its head above the parapet and exceeded the observed 'limits', with a consumption rate of 940kg/capita and GDP of US$36,972/capita. This increased cement demand resulted from growth in the construction sector driven by an increase in both public and private projects20 including the completion of the F1 site ahead of the Singapore Grand Prix.¹⁹

The post 2008 market remained relatively buoyant, with cement consumption only dipping briefly to 820kg/capita in 2010, still twice the world average for that year. This consumption was driven by an increase in public sector spending on projects including public housing and the MRT Downtown Line.¹² GDP too continued to increase despite the financial crisis, although growth was much reduced, down to just +1.1% compared to increases of 7-9% between 2004 and 2006.¹¹

The country's per capita GDP in 2010 was twice that of 2002 and resulted partly from the government's investment in economic development; an ongoing commitment that accounted for 21.1% of estimated government expenditure in 2010.¹⁴

In 2012 cement consumption reached a substantial 1035kg/capita, driven in part by a total national spend of US$20.4bn on construction. The next five years are expected to see a further increase as public sector projects including hospitals, colleges and highway developments get underway.¹³ This, coupled with a GDP/capita that exceeded US$50,000 for the first time in 2012, suggests that Singapore will remain an outlier in the GDP versus cement consumption model for at least the next few years.

UAE: For over a decade, both GDP and cement consumption in the UAE have exceeded the threshold values identified by this research. Even a significant decrease in both following the financial crisis failed to bring UAE consumption figures in line with GDP contemporaries such as the UK, which had a consumption eight times smaller than the UAE in 2010.

The trend towards this very high cement consumption prior to the financial crisis – the UAE consumed a staggering 4365kg/capita in 2008 – was driven by investments in ambitious architectural projects. These in turn drove GDP since they attracted considerable attention from both the tourism and business property sectors.¹⁵ This meant that, when the financial crisis hit, it hit hard. Construction companies that had invested heavily in the boom were suddenly unable to meet financial commitments.¹⁶

The UAE's cement consumption more than halved between 2008 and 2010. Yet the UAE remained an outlier despite this substantial hit. GDP never fell below US$34,000 and the consumption level remained above 600kg per capita. By 2012 GDP had regained pre-crisis levels thanks to government diversification and investment programmes and a young population that drove both the property and infrastructure markets.¹⁶ By contrast, cement consumption in the same year plummeted to 990kg/capita, the lowest rate in over a decade.

A recent report by the Dubai Chamber of Commerce and Industry suggests this is all set to change. The report predicts an expansion in the value of the construction industry, projecting that an increase of 0.6 million in the expatriate population will drive growth in both the residential and commercial property markets.¹⁷ This ought to ensure that the UAE maintains a position as an outlier on the graph for the foreseeable future.

Qatar: The Qatari position with regards to the GDP versus cement consumption graph has remained steadily in excess of nearly every other nation for over a decade. This is hardly surprising given that the country has the fastest GDP growth in the world²² and a low population size.

In 2003 the country's 3314kg/capita cement consumption was comparable to that of the UAE, although Qatar's GDP was far greater at US$52,414/capita. A construction boom between 2006 and 2012 continued to drive the GDP even higher and was also responsible for a staggeringly high cement consumption rate that peaked in 2008 at 4242kg/capita.

In stark contrast to the UAE, the Qatari economy remained relatively resilient during the financial downturn. Between 2008 and 2010, GDP dropped from US$84,628/capita to US$71,510/capita while cement consumption took a relatively minor hit, decreasing by approximately 10%. This buoyancy was attributed to continued government investments designed to stimulate the financial sector²² and meant that by 2012 GDP per capita had recovered to a huge US$103,900. In contrast, cement consumption in the same year decreased far more than it had even at the peak of the crisis, falling from 4254kg/capita in 2010 to 3023kg/capita in 2012.

Despite this, Qatar looks set to maintain its position as a significant global consumer of cement. A successful bid for the 2022 FIFA world cup means that there are a number of large infrastructure investments on the order books, with the government reportedly assigning 40% of its 2012 - 2016 budget to projects including a transport corridor to the capital and a new international airport.²² The Commercial Bank Capital predicts that these projects will see cement consumption in the country peak during 2013/2014.²⁰ This, coupled with a GDP that is predicted to reach US$112,000/capita by 2016,²² suggests that Qatar will continue to trend as an outlier on the GDP versus cement consumption graph.

Saudi Arabia: Unlike the other outliers discussed above, Saudi Arabia only exceeded a GDP of US$25,000/capita alongside a cement consumption rate of 600kg/capita in 2012. Having started the 10 year period from 2002 with a consumption rate three times greater than the 2002 world average at 953kg/capita; in contrast, the world average GDP for that year was barely three quarters that of Saudi Arabia's. Both measures continued to rise steadily throughout the next decade, placing Saudi Arabia firmly in the bracket of 'emerging nation' alongside contemporaries such as China and South Korea.

The blow of the financial crisis fell wide of both Saudi GDP and cement consumption, with GDP falling by US$385/capita and cement consumption falling by 103kg/capita between 2008 and 2010. Underpinned by a boom between 2004 and 2008, the entire Saudi Arabian economy fared relatively well, propelled by both government spending and economic activity throughout 2009 that promoted a strong performance in the banking sector.²³

Saudi Arabia's cement consumption further increased to 1700kg/capita in 2012, up from 1522kg/capita in 2010, while GDP rose by a staggering US$12,473/capita to US$31,800 over the same two year period.

To date, cement consumption in Saudi Arabia has been driven by investments across the board, in areas including industry, infrastructure and tourism. These projects will continue to drive consumption, with demand expected to peak in 2014.²⁴ However, with a slight decrease in cement consumption forecast for 2015 and a predicted slowing in GDP growth attributed mainly to changes in the oil sector,²⁵ it seems unlikely that both measures will remain so inflated when compared against the general global trend in the future.

China: Official and realistic figures

Another significant outlier on the graph is China. Official figures for China place the country amongst the highest consumers of cement on the planet at 1581kg/capita. While the country is clearly experiencing considerable infrastructure growth, this figure is arguably rather unrealistic given that the Chinese population in 2012 was 1.315 billion. A cement consumption of this size, in this population, would correspond to a national demand of 2.0Bnt, or around 60% total global cement production capacity.

Another factor that casts doubt on the Chinese official figures is the frequency with which even the most rapidly developing nations consume cement at a rate greater than 1000kg/capita.²⁷ Whilst not unheard of - both Saudi Arabia and Qatar have oil-fueled consumption rates far in excess of 1000kg/capita – the occurrence is relatively rare and does seem somewhat unlikely in China, where a steadily increasing GDP is still below that of the world average and far below those of both Qatar and Saudi Arabia. This throws further doubt on the health of the construction industry and the amount of capital available to drive cement demand, although, as the GDP is proportional to population size, it is perhaps not the most representative measure for dealing with a nation the size of China.

Global Cement has previously speculated that the unrealistically high official consumption rate in China results from either an overestimation by the Chinese authorities or an artificial inflation of the construction sector through unnecessary projects which lack genuine demand, or indeed a combination of both.²⁷ A more modest figure of 610kg/capita has been suggested by a peer review source.²⁶ This seems more likely as it equates to a still substantial but more realistic total national consumption of around 824Mt.

A snapshot not a long-term measure

As valuable as the GDP versus cement consumption graph is in comparing the relative economic growth of multiple nations, it is not an appropriate forecast model. This is because both measures can be independently affected by a wide number of different factors. The relationship in countries that regularly fall outside of the observed 'normal' limits of the model demonstrates how unrelated the two measures can become for some countries. Cement consumption in Singapore altered by only a small amount between 2005 and 2007 yet GDP increased by US$7813/capita over the same period. Similarly cement consumption in the UAE fell by 767kg/capita between 2010 and 2012 while GDP increased from US$34,049/capita to US$49,800/capita.

Pearson's correlation coefficient (r) is a statistical measure of the interdependence of two variables. Figure 4 shows the r values for the relationship between GDP and cement consumption in 18 countries between 2002 and 2012. A value of 0 indicates no linear relationship between the two while 1 is indicative of a perfect linear relationship (i.e. as GDP increases so does cement consumption). An r value of -1 suggests a perfect negative relationship (i.e. as GDP increases cement consumption decreases).

This analysis shows that correlation between the two measures varies significantly between nations. GDP and cement consumption are very strongly correlated in the China described by official statistics, while the two figures have a strong negative relationship in the United States. In none of the nations assessed did GDP prove to have no correlation with cement consumption. However, correlation was very low, i.e.: relatively close to 0, in five of the 18 countries investigated. These are Spain, Qatar, Japan, the UAE and Singapore.

Positively correlated (r>0)

Assuming that the Chinese statistics are inaccurate for the reasons discussed above, the world average GDP and cement consumption shows the highest positive correlation at 0.968 and indicates that the general global trend is towards increased cement consumption corresponding to increased GDP.

Vietnam: Vietnamese cement consumption has remained consistently close to the world average, increasing from 259kg/capita in 2002 to 560kg/capita in 2012. The country's GDP too has seen a significant increase over the same time period, up from US$477/capita to US$1755/capita in 2012.

Huge amounts of public capital have been invested in both infrastructure and urbanisation projects throughout the past few years, with the contribution to GDP from the construction industry increasing steadily as these projects progress.²⁸ Approximately 9 - 10% of GDP was linked to investment in sanitation, transport and telecommunication between 2004 and 2006. A number of microeconomic studies confirmed a connection between the Vietnamese growth and poverty reduction and these infrastructural investments,²⁹ effectively explaining the strong positive correlation we observed between GDP and consumption across the decade.

Brazil: Brazilian demand for cement between 2002 and 2012 tended towards a general increase, although it failed to keep pace with the average speed of growth compared to the global average. Conversely, GDP has tracked faster than the global trend, starting 2002 at slightly less than 50% of the world average and increasing to 118% by 2012.

Brazil's high correlation between GDP and cement consumption is perhaps surprising given the difference in growth rates for the two parameters under consideration in this research. Brazil's consumer driven economy kept GDP relatively buoyant throughout the financial crisis³¹ but GDP growth is now acknowledged to be slowing down.³⁰ Despite this, per capita income continues to increase with estimates suggesting that as many as 40 million citizens have risen above the poverty line since 2004.³⁰

The country's cement demand rose between 2007 and 2012 from 237kg/capita to 353kg/capita. This was driven by a successful bid for the 2014 FIFA World Cup in 2007, the Growth Acceleration Programme (PAC) initiated in the same year and preparations for the 2016 Summer Olympic Games.³²

The high correlation between cement consumption and GDP observed, then, is due to the fact that both measures are increasing. However, whether the two are actually related is debateable. Cement demand and the Brazilian economy is arguably almost entirely disassociated from GDP as the consumer products market has little direct impact on the construction industry. Infrastructure spending accounted for only 1% of Brazilian GDP in 2012.³³ Therefore, the strong positive correlation between the two measures observed in Brazil since 2002 could be coincidental. The large economic gains experienced by the country throughout the last century30 do not appear to have directly influenced cement consumption, although they have undoubtedly affected the available capital for investments in infrastructure and the bids for the major sporting events that have taken place recently.

The Brazilian construction industry is expected to grow exponentially over the next five years, placing it among the fastest growing construction markets in the world. With GDP growth already slowing and expected to decrease further,³⁰ it seems unlikely that a similar analysis to this in five or 10 years time would find the two measures to be so strongly correlated.

This highlights the main issue with using the graph as a forecasting tool. It is unable to accommodate external factors that independently influence one or both of the variables.

Saudi Arabia: Saudi Arabia's impressive GDP/capita growth means that in 2012 it exceeded the world average by 200%. In 2002 its GDP/capita was just 130% of the global average. Cement consumption growth has mirrored this trend, consistently exceeding the world average.

The Saudi economy is specialised in that the oil sector offers a large degree of flexibility that is not available to other countries.³⁴ It also has a robust tourism sector supported by two of the most popular religious shrines in the world, Mecca and Medina.³⁴ Increases in the oil export market in 2011 means that a lot of oil revenue has been earmarked for investment in the domestic economy, specifically in housing, infrastructure and economic support projects that have, in turn, driven cement demand in the Kingdom.³⁵

Negatively correlated (r<0)

United States: The relationship between GDP and cement consumption has a strong negative correlation in the US economy (-0.706). GDP/capita in 2012 was higher than 2002 while cement consumption underwent a steady decline from 2006 onwards.

The country's economy is based on a service model, which meant that the financial crisis had a significant impact. GDP/capita remained flat between 2007 and 2010 but cement consumption has plummeted since 2008. The construction industry in the US underwent a severe contraction in response to the financial crisis. The result was reductions in both residential and commercial projects, state budgets and public works. Cement demand fell in response, with total consumption figures in 2010 nearly half the level of just five years before.³⁸

The Portland Cement Association (PCA) suggests that 2013 was the first year in which this trend began to reverse, with real construction spending expected to increase further in 2014. The PCA also predicted that cement consumption will regain pre-crisis levels by 2018,³⁶ suggesting that the correlation between the two measures will not remain so markedly negative in future analyses.

South Korea: The negative correlation between the two measures is not as dominant in South Korea as the US at -0.344 and is likely the result of a construction industry that has been flailing since 2008. Indeed, cement consumption in South Korea has been particularly changeable. In 2002, the value was firmly above the 1000kg/capita mark at 1216kg/capita but had fallen to 955kg/capita by 2005. This trend has been sustained throughout the decade as consumption rates flirted with 1000kg/capita until 2010, when a demand of 950kg/capita marked the beginning of a decline. GDP also decreased significantly between 2010 and 2012 but, unlike cement consumption, had risen steadily until this point.

The construction industry in South Korea has been struggling since 2008 due in part to poor property sales and increasingly tight lending conditions.³⁹ This slump has been identified as a particular challenge to the industry's growth, with analysts looking to the 2018 Winter Olympics to offset the trend.⁴¹

South Korea operates a market economy, one with a growth rate limited by both labour regulations and a rapidly ageing population. With a strong focus on exports (they account for nearly half the country's GDP), South Korea was hit hard by the global financial crisis but recovered within a matter of years.⁴²

It is easy to see then why GDP and cement consumption would have such a low correlation in this example. There is a struggling construction industry and an economy that is reliant on exports and subsequently market trends in other nations.

However, the correlation between the two may not remain negative for long. With the US – Korea Free Trade Agreement of 2011 expected to increase the value of the export market by billions of dollars year-on-year and the 2018 Winter Olympic Games set to regenerate the construction industry,⁴¹ future analysis along the lines of this study would likely find that increases in both correspond to a more positive correlation; whether that correlation is the result of a direct relationship between the two measures or not.

Spain: The correlation between GDP and cement consumption in Spain, although negative, is also very weak at only -0.143. This is not unpredictable given the huge changes that the country has witnessed since its 2007 crash. An analysis of GDP and cement consumption 10 years ago would likely have found the two factors to be very strongly correlated. The Spanish construction industry and economy were both booming prior to 2007. The construction industry grew quickly from 1999, registering 11% growth in demand in 20005 and a further 5.1% increase in 2001. This growth was sustained and promoted a year-on-year increase in cement demand that peaked at 1300kg/capita in 2007, placing Spain among improbable contemporaries such as South Korea and Saudi Arabia on some earlier consumption graphs. GDP too increased quickly during this time, nearly doubling between 2002 and 2007.

A difficult 2009 saw a substantial constriction of the building industry, with a 60% fall in total residential construction projects compared to 2008 levels. The result was a 32.9% decrease in cement consumption in 2009 followed by a further 15.2% in 2010.^7-8 This meant that by 2012 Spanish consumption had decreased from 936kg/capita in 2008 to 438kg/capita.

Spain's GDP followed a similar trend, nearly doubling between 2002 and 2007 before declining. However, exports remained relatively resilient throughout this time and the country's mixed economy meant that the decline was not as marked in GDP as in cement consumption.⁴² This undoubtedly contributes to disassociation between the parameters.

The two measures do not look set to become correllated (either positively or negatively) in the near future. Oficemen predicts a continued contraction in the construction industry and noted a further decrease in per capita cement consumption in 20132 while a modest economic growth was recorded for the same year.⁴²

The graphs' value as a forecast tool

The correlation between GDP and cement consumption across multiple nations and years demonstrates that both are liable to be influenced by external factors, themselves subject to variation.

GDP, as the market value of goods and services, can be hugely impacted by factors including war, oil incomes and other exports. In contrast, cement consumption is a reflection of the demand for housing, infrastructure and is related to a country's level of urbanisation. It is driven by government investments, bids for sporting events and large-scale private sector projects in some nations, for example Saudi Arabia, the UAE and Qatar.

It is clear that GDP/capita and cement consumption per capita do not operate in sole conjunction with the other. There are many situations in which growth in GDP facilitates an increase in cement consumption; the strong correlation between the two in emerging nations like China and Vietnam are evidence of this. Infrastructural investments in these nations are necessary to facilitate increases in industrialisation, which in turn increase GDP and facilitate progress. Likewise, there are situations in which the growth in GDP has no bearing on consumption, for example trends in the oil sector have a huge bearing on Saudi GDP but little or no effect on cement consumption.

This disassociation under pressure from external factors arguably makes the use of the relationship between the two measures an inappropriate forecasting tool, especially since their correlation varies between countries. The current analysis shows that, even within individual countries, the correlation coefficient is liable to change over time. However, the mean global values are very strongly correlated and, providing that the graph is not used to forecast likely trends in either, it still offers a good comparison of the relative economic position between different nations at a set point in time.

References

Adisa, O.K, 2013. Economy of RHA (Rice Husk Ash) in Concrete for Low-Cost Housing Delivery in Nigeria. Journal of Civil Engineering and Architecture, 7(11). Pp 1464-1470

Aggregate Research, 2012. S. Korea's cement exports hit all-time high. Aggregate Research [online].

Aijazira Capital, 2011. Saudi Cement Sector. [Online].

Al-Ansay, M. 2010. Innovative Solutions for Sulphur in Qatar. Qatar Shell Research and Technology Center. [online].

Al-Hamidy, A., 2009. The global financial crisis: impact on Saudi Arabia. Saudi Arabian Monetary Agency [online].

APCMA, 2012. Pakistan Cement Sector. [online].

Armstrong, T. 2013. An overview of global boards cement sector trends. [online]

British Geology Survey, 2008. Cement. Mineral Planning Factsheet. [online].

Building and Construction Authority, 2012. Public Sector projects to sustain construction demand in 2012. [online].

CEMBUREAU, 1999. 'Best Available Techniques' For The Cement Industry. [Online].

CEMBUREAU, 2000. Activity Report 2000. CEMBUEARU[online].

CEMBUREAU, 2008. Activity Report 2008. [online].

CEMBUREAU, 2009. Activity Report 2009. [online].

CEMBUREAU, 2010. Activity Report 2000. [online]

Cemnet, 2013. Cement demand in Belgium falls by 2.2% while imports rise. [Online].

Cemnet, 2013. CIS cement developments. [online]

Chen, W. 2006. Hydration of Slag Cement: Theory, Modeling and Application. [Online].

Collantine, K., 2008. Thoughts on Singapore's F1 Street Track (2008 Singapore Grand Prix Preview). F1 Fanatic, the Formula 1 Blog. [online].

Commercial Bank Capital Sector Report, 2012. Qatar Construction Sector. [online]

DesignBoom, 2008. The Future in Dubia. [online].

Dubai Chamber of Commerce and Industry, 2014. Construction Sector to Contribute 11.1% to UAE's GDP in 2015. [Online].

The Energy Resource Institute, 2007. Teri Energy Data Directory and Yearbook, 2005/06. TERI Press, New Delhi .[online]

The Economic Development Sector, 2010. Expenditure Overview. [online].

Economic Research Department Business Development Division, 2010. Housing and Real Estate Market Development –Worldwide Experience and Options for Vietnam. The International Conference 2010. [online].

Edwards, P., 2012. Cement in the USA. Global Cement Magazine [online]

Edwards, P. 2012. Massive Cement growth ahead for Qatar. Global Cement Magazine [online]

Edwards, P., 2013. China: First in cement. Global cement magazine. [online]

EUbusiness, 2014. Spain: country overview. EUbusiness [online]

European Ready Mixed Concrete Organization (ERMCO), 2013. Ready Mixed Concrete Industry Statistics Year 2012. [Online].

Ficem, 2013. Statistical Review. Ficem [online].

Heildberg, 2012. Annual Report. Heildberg [online].

Holcim, 2013. Holcim Romania: Professionally Competent [online].

Index Mundi, not dated. Qatar GDP. [online].Oficemen, 2010. Cement Consumption will fall 15% in Spain in 2010. [online]

Le-Hoai, L., Dai Lee, Y. and Yong Lee, J. 2008. Delay and cost overruns in Vietnam large construction projects: a comparison with other wselected countries. KSCE Journal of Civil Engineering. 12(6). pp. 367-377 [online].

Lynn and Pancholia, 2008. UAE Cement. Al Mal Capital. [online].

McArdle, M., 2014. The miracle behind Brazil's economic miracle. Boomerang View [online].

Mehta, R. and D'souza, A., 2008. Cement, Annual Review. CRISIL Research.

The Ministry of Trade and Industry, 2008. Press Release: GDP Growth to Moderate to Medium Term Potential. [online].

The Ministry of Trade and Industry, 2009. Press Release.

NCB, 2012. Comparative Energy Costs and Expansionary Fiscal Policies are Fuelling the Saudi Cement Sector. Saudi Cement Sector Review. [online].

Monaco Statistics, 2012. Gross Domestic Product. [online].

Oficemen, 2014. The public investment slump drags down the consumption of cement at levels of 50 years ago. Oficemen [online]

Oxford Business Group, 2011. Economic update. Saudi Arabia: economic development. Oxford Business Group [online].

The Portland Cement Association, 2013. U.S. cement consumption to reach nearly 80 million tons in 2013. [online]

Railway Technology, 2013. Downtown Line Stage 3, Singapore. [online].

Rapoza, K., 2012. Brazil economy not what it once was. Forbes [online].

Research and Markets, 2012. UAE Construction industry May 2012. [online].

Schaller, S., 2008. Lula's growth acceleration program: the best that Brazilian government finding can buy? Council on Hemispheric Affairs. [online].

SEMAPA, 2012. Annual Report. [online]

Susilo, E., 2008. Singapore-Country Report. The 15th Aisia Construct Conference. [online].

Timetric, 2013. Travel construction in South Korea –key trends and opportunities to 2017. Timetric [online].

Tuan Anh, N., and Filoni, E. 2007. Case studies and Experiences: Vietnam's Emerging Cement Industry. Emerging Market Reports [online].

USG, 2010. Minerals Yearbook 2008.

Ventures Middle East, 2011. The Saudi construction industry. Ventures Middle East [online].

Vicat, 2013. Annual Report [online].

The World Bank. Population data for all countries and years. [Online].