Why Agricultural Infrastructure Matters

The last decade saw unprecedented spikes in food prices. Food riots in the developing world became an all too common occurrence as market volatility spread throughout food chains and pushed prices out of the reach of the average consumer.

Export led growth can give countries the hard currencies needed to buffer the shock of food prices. But how do countries increase their share of exports? The consensus would seem to state that a strong emphasis on a mix of infrastructure, property rights, government spending, and business and trade freedom would yield the best results. However, developing countries have to make hard choices of where to target their funds to get the best results. Which of these will have the greatest return on investment, dollar for dollar?

Our staff member, recently explored this question and has allowed us to publish her findings in detail below. She found that, dollar for dollar, a developing country should focus on infrastructure first, whereas more developed economies may want to focus on other factors. Therefore, investing in agricultural infrastructure – networks such as Agroam – has a two-fold effect on the supply chain: the volume of agricultural products that can enter the market, greatly reducing the volatility of markets. For when that isn’t enough, the additional hard currency generated by agricultural exports can help support farmers and consumers when it’s needed.

The case is clear, agricultural infrastructure is the most cost effective way to strengthen local markets – and Agroam is happy to be on the forefront of providing this technology. The full report can be found below:

Explaining Export Growth: Differences Between Least Developed Countries and the Rest of the World

Literature Review

International trade has become increasingly global over the last decades due to wholesale changes in communication and transport technology. In search for the most competitive offers, stakeholders progressively source products from an ever larger variety of countries. Globalization has been heralded as one of the key factors of economic growth on a country-wide level (Behar and Venables 2010). The Global Growth Report (2010) finds that while it is important to increase competitiveness within domestic markets, growth can only be maximised when countries are also seeking to export goods.  Thus, the promotion of exports is crucial to outward-looking growth strategies, which is associated with job creation and foreign exchange earnings (Global Growth Report 2010). To be able to benefit from emerging export prospects, countries need to provide an environment that enables trade facilitation and that ensures to attract global demand (WEF 2010). This study seeks to understand which factors play the biggest part in explaining export growth.

In order to determine factors that are believed to contribute to export growth, a literature review is executed and finds evidence for the existence of six aspects that are said to contribute to export growth. The first aspect concerns the degree to which a country is open to international trade. For the purpose of this study trade openness is defined in the simplest terms as absence of trade tariffs, import quotas and export subsidies (David 2007). Scholars voice divided opinions on whether open economies benefit from greater export potential or not. Scholars of the neo-liberal school of thought advocate the mutual benefits of openness to imports and exports (Krueger 1997, Edwards 1992). Whereas Yanikka (2002) argues that especially economies based on primary commodities should levy export subsidies in order to diversify revenue streams.

Secondly, the scale of government investment is believed to influence export growth. Again, there is no consensus on whether government spending has a positive or negative effect on export growth. In order for countries to gain from international commerce, governmental investments need to be put in place to strengthen exports. Hays et al (2005) argue that if government spending is used to reimburse domestic agents for lost revenues due to exports, government spending has a negative effect on export growth. Government spending within low-income and resource-dependent countries has been a topic that has generated particular scholarly interest. Davis and Tilton (2005) find that low-income countries that are resource rich can significantly enhance economic prospects, and exports, when rents generated are re-invested in policies that seek to steer the economy away from a dependency on primary commodity exports.

In addition, it is argued that the degree of general “business-friendliness” of a country contributes to explaining export growth. There is large consensus among the literature that the enactment of business-friendly policies contributes positively to export growth (Djankov et al 2006, Barro 1996, Hall et al 1999). Large consensus among scholar is also evident on the role that property rights play in accelerating export growth. Anderson and van Wincoop (2004) argue that built-in trade costs, such as weak property rights, are often more important to impeding exports than direct tariffs on goods of third parties. This is further substantiated by Maskus’ and Penubarti’s (1995) finding that “increasing patent protection has a positive impact on bilateral manufacturing imports into both small and large developing economies”.

Consensus largely suggests that infrastructure, especially transport, can be viewed as the single most important factor that influences export and trading patterns (Baldwin 2012). The report “Enabling Trade: Value Growth Opportunity” by the World Economic Forum in partnership with the World Bank estimates that investments into more efficient transport is among the most important measures to improve gains from international trade. Behar and Venables (2010) argue that transport costs and volatility in transport prices affect the ability of a country to process exports. The uncertainty and additional costs associated with inefficient transport services are thus regarded as the main obstacle to greater exports. Hoekman and Naticin (2011) find that the overall logistics performance of a country is one of the most important measures that can either enable or obstruct maximum exports. Most non-industrialized countries suffer from weak logistics capacity, impeding economic growth.  It is thus crucial that less developed countries pursue policies that further strengthen their logistical capacity.

Gross Domestic Product (GDP) has been regarded to symbolize the relative health of an economy compared to other economies by measuring economic production. For the purpose of this study, GDP is treated as a proxy for social welfare and progress. Scholars agree that there is large correlation between the size and volume of GDP compared to its export production. Thus, it is assumed that the relative size of the GDP indicates whether a country is fairly developed economically (large GDP) or trailing behind (small GDP).

This study seeks to fill the research gap by testing whether it is true that aforementioned sectors significantly impact variance in exports, or whether further research would need to be done in order to identify variables that are more salient in explaining variance in exports. In addition, the study attempts to clarify in whether there are differences between drivers of export when looking at different development stages of country. If there should be meaningful differences between variables explaining export growth in LDCs and non-LDCs, further research would have substantiate in how far this has implications for economic policies.

Study Design

Participants

The study focuses on 155 individual countries worldwide. These countries are diverse in their nature when it comes to their infrastructure capacity, business friendliness, legal framework, GDP and governmental spending. High-Income countries such as the United States as well as Switzerland are included in the sample and generally have greater capacities than less wealthy countries. On the other hand, the sample also includes least developed countries, such as Afghanistan or Zambia, in which people are surviving on one US dollar a day and who lack various institutional capacities (e.g. infrastructure).

 

Measure

The Heritage Foundation computes the “Economic Freedom Index” which consists of ten subcomponents and is ranked from 0-100, with 0 being the lowest and 100 being the highest score possible. Four of our independent variables are taken from the Economic Freedom Index.  The subcomponent property rights is taken to represent the variable legal framework, which measures how far a country’s laws guarantee property rights and the degree to which governments enforce compliance with such laws. The subcomponent government spending is used to represent the variable government investment. It takes the ratio of governmental spending to GDP as a measure, with 0 indicating high governmental spending and 100 indicating low government spending. The subcomponent business freedom is used to represent the variable business friendliness. The business freedom component is an overall indicator on how easy it is to conduct business in a particular country and takes into account how simple it is to open and close a business in a country. The subcomponent trade freedom is employed to represent the variable international trade capacity. The subcomponent measures in how far tariffs and non-tariff measures are absent and scores a country accordingly. GDP is represented in terms of purchasing power parity (PPP). The expenditure measure of calculating GDP takes into account consumption, government spending, investment and net exports. Looking at GDP in terms of PPP eliminates the effect that different price levels and changes might have and is therefore better equipped to illustrate the output of an economy when compared to other economies (World Bank 2013a).

The Logistics Performance Index (LPI) computed by the World Bank aims at measuring individual logistics performances on a country basis worldwide. It employs six subcomponents that individually measure different aspects of logistics performance, such as infrastructure and customs, and ranks performance from 1-5, with 1 being the lowest and 5 the highest score obtainable. This study employs the subcomponent infrastructure, because we are not interested in the overall logistics capacity but very specifically in the role of infrastructure. Because other independent variables, except GDP, are ranked form 0-100, the variable infrastructure was transformed to reflect a scale of 0-100, in order to facilitate comparisons.

Data on the dependent variable, exports, is compiled by the United Nations Conference on Trade and Development (UNCTAD), which reports a country’s exports in accordance with the recommendations of UN International Merchandise Trade Statistics in terms of US Dollars (USD) (UNCTAD 2013).

Method

Having screened the data, any inadmissible values (such as gaps in the data appearing as 0) were re-coded as missing. Cases with missing values are excluded listwise, so that countries lacking a single piece of information are excluded entirely. It is believed that the representativeness of the sample will not be critically comprised in doing so, because the sample consists of data on almost every country, thus it yields a high degree of representation regardless. Countries are coded into LDCs, as identified by the World Bank, and non-LDCs. Thus, our sample can be split into two different groups, when interested in differences between LDC and non-LDC, or treated as one group, when interested in worldwide effects.

It is firstly expected to find that infrastructure is going to be the most important variable determining export growth. Secondly, it is expected to find that there are differences between LDCs and non-LDCs when it comes to infrastructure capacity, rule of law, business freedom, government spending and GDP. Nonetheless, it is thirdly expected to find that regardless of differences between LDC and non-LDC, infrastructure will be the most important variable in explaining variance in exports.

Thus, the following three hypotheses are being tested:

H1: Infrastructure is the most important variable determining variance in exports

H2: There will be significant differences between LDCs and non-LDCs in terms of our independent and dependent variables

H3: Infrastructure remains the most important variable in explaining export variance when controlling for LDCs and non-LDCs respectively

The hypotheses are tested with the following equation:

exports=  α+(bproperty rights + bgovernment spending+ bbusiness freedom+ btrade

freedom+ binfrastructure + bgpd) + ε

Descriptives

 

The group LDC displays lower means for every variable when compared to the non-LDC group, except for government spending. Limited government spending is ranked highly by the Heritage Foundation, i.e. governments of LDCs spent less (n=30; mean= 76,31; s=18.48) when compared to non-LDCs (n=121; mean=62,83; s=23.50). Both groups differ the greatest when it comes to property rights and infrastructure. The LDC group (n=31) averaged a score of 27,67 (s=10,81) for property rights, and obtained an average score of 40,02 (s=6,48) for infrastructure (n= 34). In contrast, the non-LDC group (n= 121) averaged a score of 48,39 (s=25,29) for property rights and obtained an average score of  56,32 for infrastructure (n=121, s=14.30).  Differences in terms of trade freedom and business freedom are smaller. Property rights obtained the lowest score in both groups with infrastructure receiving the second lowest score within both groups. Thus, both groups exhibit similar patterns of the relative distribution of scores, however, the non-LDC group scores higher in absolute terms. The average GDP of the group of non-LDC is roughly 20 times higher than that of LDCs.

Correlation

In order to establish whether it is feasible to run a general linear model (GLM), we test whether the independent variables amongst each other are related. In order to do so, bivariate correlations are executed. It is found that there is a significant relationship between all of the independent variables (p<.001). All variables are positively related with each other with at least r=.48, or higher. The greatest correlation can be observed between property rights and infrastructure with r=.8.

Generally, property rights is strongly correlated with all other variables. The weakest correlation can be observed between trade freedom and GDP, r=.48. Government spending is negatively related with all other variables, the weakest correlation being r=-.3, between government spending and trade freedom, and the strongest r=-.40, between government spending and property rights. Government spending is negatively correlated with all other variables because the Heritage Foundation ranks least government spending highly and extensive government spending lowly. Thus, high government spending is positively correlated with the remaining variables. Having established that all variables are reasonably correlated with each other, a GLM will be executed in order to assess which variable best predicts export growth and whether differences between the two groups, LDC and non-LDC can be observed.

Independent Variables

1

2

3

4

5

6

1   infrastructure

1

2    property rights

,797**

1

3    government spending

-,332**

-,403**

1

4     business freedom

,637**

,731**

-,379**

1

5     trade freedom

,540**

,530**

-,306**

,524**

1

6     GDP

,786**

,770**

-,357**

,560**

,476**

1

**. Correlation is significant at the 0.01 level (2-tailed).
b. Listwise N=141

General Linear Model- Country

In order to test the first hypothesis, in which infrastructure is the most important variable in explaining variance in exports, independent of country, a general linear model is run.  Results indicate that while the overall model is significant (F(151, 1)=13.32, p<.05), the type of country had no effect on export growth (p>.05).  The entire model explained 38% of variance within the dependent variable export. (adjusted R²= .38). Given the fact that our total sample comprises proportionally less countries that are in the group of LDC (n=27) than in the group of non-LDC (n=124), it might be the case that the main effect of country is not being picked upon because our sample for LDC is more than four times smaller than the non-LDC group. However, infrastructure has a significant (p<.05) effect on exports with η=.315. GDP fell slightly short of being significant with p=.053. However, its effect size is very small with η=.03, especially when compared to the effect size of infrastructure, as illustrated with table 3. Thus, the hypothesis that infrastructure is the most influential variable that causes export growth has been accepted within the overall model.

Because the ANOVA testing for between-group differences found that all means were significantly different between both groups, it is decided to run another GLM that controls for LDCs and one that controls for non-LDCs. Both GLMs are computed in order to test whether there is truly no difference between what causes export growth in LDCs and non-LDCs, as initially indicated by our first model, or whether there is differences between the both groups. It is crucial to further assess whether there is significant differences between the both groups, because if there are no differences between both groups than policy recommendation do not need to be tailored for each group specifically. In the case that there are differences between both groups, policy advice can be targeted more efficiently towards LDCs and non-LDCs.

 

Table 3 Tests of Between-Subjects Effects- Country

Source

F

Sig.

Partial η²

Corrected Model

15,124

0,00

0,407

Intercept

24,328

0,00

0,156

Country

0,512

0,475

0,004

infrastructure

60,635

0,00

0,315

GDP

3,815

0,053

0,028

property rights

2,982

0,087

0,022

government spending

0,042

0,838

0,00

business freedom

0,122

0,727

0,001

R Squared = ,407 (Adjusted R Squared = ,380)
Dependent Variable:   exports

 

General Linear Model- Non-LDC

The second hypothesis, that infrastructure remains the most important variable no matter which category the country belongs to, is tested by running a GLM controlling for LDCs. When running a GLM controlling for LDCs a different result is produced than when including LDCs.  For the group non-LDCs there is a significant main effect of infrastructure and GDP (p<.05). Infrastructure has the greatest effect on export growth with η=.38. Compared to the initial model that is indiscriminate towards the type of country, the model controlling for LDCs explains roughly 40.1% of variance in exports, which is a modest boost of 2.7%.  This can be explained by the fact that countries that have stringent property rights in place, a business-friendly environment, freedom of trade and exercise appropriate amounts of government spending, are left to invest in their infrastructure in order to boost export growth. GDP only has a small effect size with η=.05. Thus, it becomes evident that it matters less how wealthy a country is compared to other countries, as long as a certain standard of wealth is achieved already.

Table 4 Tests of Between-Subjects Effects, non-LDCs

Source

F

Sig.

Partial η²

Corrected Model

13,801

0,00

0,441

Intercept

7,136

0,01

0,064

infrastructure

64,192

0

0,379

property rights

3,49

0,07

0,032

government spending

0,294

0,589

0,003

business freedom

0,018

0,892

0

trade freedom

2,248

0,137

0,02

GDP

5,003

0,027

0,045

a. R Squared = ,441 (Adjusted R Squared = ,409)
Dependent Variable:   exports

 

General Linear Model- LDCs

In order to further corroborate the hypothesis that infrastructure remains the most important variable in explaining exports, we control for the group non-LDCs and run a GLM for the group LDCs only.  In contrast to the model accounting for non-LDCs, the GLM model for LDCs countries yield different results. The overall model is significant (p<.05) and explains 80.3% (adjusted R²= .803) of variance, which is roughly twice as much as our previous model does.

When controlling for the group non-LDCs, property rights and GDP become significant in explaining variance in export (p<.05). GDP is the most powerful variable explaining exports (η=.83). This is in stark contrast to the previous model in which GDP explained only a small amount of export growth. In addition, property rights also become significant when explaining export growth among LDC (η=.43).

Source

F

Sig.

Partial η²

Corrected Model

18,69

0,00

0,849

Intercept

1,412

0,25

0,066

infrastructure

1,158

0,295

0,055

property rights

14,903

0,00

0,427

government spending

2,841

0,107

0,124

business freedom

0,519

0,48

0,025

tradef reedom

0,143

0,709

0,01

GDP

95,179

0,00

0,826

a. R Squared = ,849 (Adjusted R Squared = ,803)
Dependent Variable:  exports

 

Discussion of Results

This study tests the hypothesis which states infrastructure is the most important variable when determining variance in exports, when all countries are entered into the model. Thus, it is illustrated that overall infrastructure is among the most important factor which causes variance in export, regardless of what type of country is looked at. This has been largely predicted by our literature review, in which it was found that there is large scholarly consent on the salience of infrastructure. It is hard to argue with the realities that in order to transport goods, roads, ports and efficient bureaucracies need to be in place to process exports.  Furthermore, the second hypothesis was also confirmed in that our model established that there are significant differences between LDCs and non-LDCs. Scholars have long advocated that LDCs face particular problems when it comes to growth of exports. LDCs can be largely classified as small-island states or land-locked countries. Both geo-political situations are a hindrance towards the smooth administration of exports, as they face hurdles that, e.g. a mainland-based country with a coastal line does not face. Interestingly, it was found that other variables such as business freedom, trade freedom and government spending only have an explanatory capacity when in conjunction with other variables, such as property rights, GDP or infrastructure. This lends credence to the assumption that there is qualitatively difference between the first group of variables, that cannot explain variance in exports individually, and those variables that can explain variance on an individual basis. It is argued that institutions such as robust property rights and infrastructure are the very prerequisite to any form of trade. Legal rights, such as intellectual property rights, secure investments made and thus encourage entrepreneurial activity, whereas infrastructure enables the physical trade of goods. Trade orientations, government spending and business-friendliness are only of secondary concern to the primary necessities of legal institutions and physical capacity.

However, the third hypothesis that infrastructure remains the most important variable when controlling for LDCs and non-LDCs has only been partially accepted. It was found that for non-LDCs, infrastructure yields the greatest power in understanding variance in exports, with GDP being able to only explain a small amount of variance in exports. The model controlling for non-LDCs found that GDP generates the greatest importance in explaining variance in exports, with property rights being the second most important item. Surprisingly, it was found that infrastructure is not significant in explaining variance in exports when controlling for non-LDCs. It is suggested that the most likely explanation for the importance of GDP in explaining variance in exports lies in the fact that a small number of LDCs, such as Angola or Sudan, export a relatively large of amount of extractive primary commodities to the rest of the world. This would also explain why infrastructure is not significant in explaining export growth among LDCs. Often, there is just about enough infrastructures in place for LDCs to manage to extract primary commodities and transport it to the next harbour. Thus, it is more crucial to have property rights in place that protect the direct investment of foreign oil companies than to expand infrastructure on a country-wide level.

Further research is required to establish whether the above mentioned hypothesis holds true when it comes to the importance of GDP in exports of LDCs.

Bibliography

Abe, K & J Wilson (2009), ‘Weathering the Storm: Investing in Port Infrastructure to

Lower Trade Costs in East Asia’, World Bank Policy Research Working Paper 4911.

Anderson, J & van E Wincoop (2004), ‘Trade costs’, Journal of Economic Literature, 42

(3), 691-751.

Arvis, Jean-François, Monica AlinaMustra, LauriOjala, Ben Shepherd, and Daniel Saslavsky. 2010. Connecting to Compete 2010: Trade Logistics in the Global Economy. Washington, DC: World Bank. World Bank, Logistics Performance Index- Connecting to Compete 2012

Arvis, Jean-François, Monica AlinaMustra, LauriOjala, Ben Shepherd, and Daniel Saslavsky. 2010. Connecting to Compete 2010: Trade Logistics in the Global Economy. Washington, DC: World Bank. World Bank, Logistics Performance Index- Connecting to Compete 2010

Baier, S & J Bergstrand (2001), ‘The growth of world trade: tariffs, transport costs, and

income similarity’, Journal of International Economics, 53, 1–27.

Baldwin, R (2012), “WTO 2.0: Global governance of supply-chain trade”, CEPR Policy Insight, 64, December.

Barro, Robert J, 1996, ”Determinants of Economic Growth: a Cross-Country Empirical Study”,

NBER Working paper 5698.

Behar, A, P Manners & B Nelson (2009), ‘Exports and Logistics’, Oxford Department of

Economics Discussion Paper 439.

Commission on Growth and Devleopment. 2010, Global Growth Report: Strategies for Sustained Growth and Inclusive Development, accessed 07.05.2013 http://www.ycsg.yale.edu/center/forms/growthReport.pdf

David, H.L. (2007),  A Guide to Measures of Trade Openness and Policy, Indiana University of South Bend, http://www.cgu.edu/Include/spe/Conf%20Pol%20Econ%20Indicators%202007/A%20Guide%20To%20Measures%20of%20Trade%20Openness%20and%20Policy.pdf (access date May 15, 2013).

Davis, S.; Tilton, J.E. (2005), The resource curse, Natural Resources Forum 29 (2005) 233–242

Djankov, S.; McLiesh, C.; Ramalho,R. (2005), Regulation and Growth, The World Bank, http://rru.worldbank.org/documents/discussions/regulationofgrowth.pdf (access May 20, 2013).

S. Edwards. Trade orientation, distortions and growth in developing countries. Journal of Development Economics, 39(1):31-57, 1992.

Francois, J, K Kepler& M Manchin (2007), ‘Institutions, Infrastructure, and Trade’,

World Bank Policy Research Working Paper 4152.

Guilherme, José and Reis; Farole,Thomas, ” Trade Competitiveness Diagnostic Toolkit. World Bank”, Washington, DC

Hall, R. E.; Jones,C.E. (1999), ”Why do some countries produce so much more output per worker than other”, Quarterly Journal of Economics 144 (1), 83-116

Hays, J.C.;Ehrlich, S. D.; Peinhardt, D. (2005), Government Spending and Public Support for Trade in the OECD: An Empirical Test of the Embedded Liberalism Thesis, International Organization, 59(2): 473- 494

Hoekman, Bernard, and Alessandro Nicita. 2011. “Trade Policy, Trade Costs, and Developing Country Trade.” World Development 39 (12): 2069–79.

Hoekman, Bernard and Jackson, Selina (2013) Reinvigorating Trade Policies: Think Supply Chain!, accessed 07.05.2013,  available at http://www.voxeu.org/article/reinvigorating-trade-policy-agenda-think-supply-chain

Krueger, A. Trade policy and economic development: How we learn. American Economic Review, 87(1):1-22, 1997.

Kunaka, Charles; Monica,AlinaMustra, and Sebastian Saez. 2012. “Trade Dimension of Logistics Services.” World Bank, Washington, DC

Maskus, K.E.; Penubarti, M.  (1995), How trade-related are intellectual property rights?, Journal of International Economics, 39(3-4): 227- 248

UNCTAD (2013), Methodology and Classification, http://unctadstat.unctad.org/UnctadStatMetadata/Classifications/Methodology&Classifications.html#EXPANATION OF SYMBOLS (access May 05, 2013).

World Bank(2013), Region- Least Developed Countries, http://data.worldbank.org/region/LDC (access May 05, 2013).

World Bank (2013a), GDP PPP based, http://data.worldbank.org/data-catalog/GDP-PPP-based-table

(access May 05, 2013).

World Economic Forum, Bain & Co. and World Bank (2013), Enabling Trade: Valuing Growth Opportunities.