Mass Media and Food Crisis: A Discourse Analysis of Media

Hatice Yuksel ˡ , *, Konstantinos Karantininis ², Sebastian Hess ³

ˡMediterranean Agronomic Institute of Chania (MAICh), Greece

² Swedish University of Agricultural Sciences (SLU), Sweden

³ Swedish University of Agricultural Sciences (SLU), Sweden

*Corresponding author. Tel: +90-507-178-9353E-mail address: yukselhtc@gmail.com (Yuksel, H.)

Paper to be presented at the ECPR General Conference, Bordeaux, 4-7 September 2013.Panel P400: Value Conflicts and Diverse Claims in Food Governance across Scales.

2

Mass Media and Food Crisis: A Discourse Analysis of Media

Abstract

In 2007-2008, when food prices started to increase dramatically, purchasing power parity of

consumers, especially the urban poor, started to decrease automatically. High food prices were

argued to cause poverty, hunger, and food riots among urban populations. Henceforward, ‘food

crisis’ became a new story line on the current debate. In contrast, in the pre-2007 period, when

rural farmers had been facing negative welfare effects of low food prices for many years, there

were no crisis talks. This paper analyzes media’s shifting discourses for the 2000-2013 period

and propounds media bias on the food crisis debate by using content analysis and OLS regression

model. Further research can explore how media bias translates into policy bias.

Keywords: food crisis; media bias; content analysis; relative incomes

3

1. Introduction

"We have all become addicted to breaking news on a crisis like this."

Eli Flournoy, director of CNN's international news source

In 2007-08, world food prices reached record levels, rising 80% in 18 months. Following

this peak, food prices fell, but since 2009 the cost of food has been climbing steadily in global

markets, reaching record highs again in 2011. Over the last five years, the FAO food price index

has risen by 92%, and many observers expect food prices to continue to rise, threatening the lives

and livelihoods of millions of people (FAO, 2012: 2).

A wide range of research has examined the (potential) impacts of higher food prices on

poverty (Ahmed et al., 2007; Aksoy & Isik-Dikmelik, 2010; Dessus et al., 2008; Headey, & Fan,

2008; IMF, 2008; Ivanic & Martin, 2008; Wodon et al., 2008; Wodon & Zaman, 2008; World

Bank, 2008; Zezza et al., 2008). Although we still do not know the accurate welfare effect of

higher food prices on urban, rural and country level poverty, the findings of the studies suggest

that the overall impact of higher food prices on poverty is generally adverse. For some studies

this conclusion is much more obvious for urban consumers. Surprisingly, the study by Ivanic and

Martin (2008) shows that rural poverty increases more than urban poverty does, for two out of

three countries examined. However Aksoy and Isık-Dikmelik (2010) analyze some of the same

surveys as Ivanic and Martin (2008), and they end up with different results. The real welfare

effect of higher food prices depends on whether the poor are net buyers or net sellers of food and

whether they are marginal net buyers or sellers of food, since rural populations also have large

numbers of net buyers of food.

It is well known that low prices on the world market benefit net buyers and hurt net sellers

of the specific product. If net buyers of food are hurt by recent high food prices, net sellers of

food should have been hurt by low food prices in the years prior to 2007. This principle is well

known; however, we do not find its implications reflected in media arguments. Since public

officials react to media news because they see it as a reflection of public opinion (Kim, 2005: 3),

the media are important actors in policy making. Paying a disproportionate amount of attention to

4

the problems of urban consumers and yet not paying any attention to the problems of rural

farmers (i.e. media bias) will translate into bad policies (i.e. policy bias).

A convenient approach to analyze the media’s arguments is discourse method, which has

been taken up in a variety of social science disciplines, including linguistics, sociology,

anthropology, psychology, international relations, communication studies, and has recently

become an important method in economics literature. The present paper proposes content

analysis (CA) as a complement to discourse analysis (DA). It then uses the data derived from CA

for further statistical analysis. CA is a research technique for the objective, systematic, and

quantitative description of the manifest content of communication (Berelson, 1952: 18). It is,

then, a quantitative method applicaple to what has traditionally been called qualitative material –

written language (Zito, 1975: 27). Quantitative CA is replicable examination of symbols of

communication, which have been assigned numeric values according to valid mesurament rules

and the analysis of relationship involving those values using statistical methods, to describe

communication, draw inferences about its meaning, or infer from the communication to its

context (Fico et al., 2005 :25).

The paper starts by presenting a simple theoretical framework to appraise the possible

welfare effects of food price changes. It then moves to presenting the main discourses of

international media on food crisis debate. It presents the methods which are used and

demonstrates the results. It then discusses whether the media discourses are biased, and if so

which factors might lead to this media bias, and whether media bias translates into policy bias.

Finally, it draws conclusions.

The present study seeks to provide evidence on the questions arising from the dramatic

increase in food prices. Are media biased on the food crisis debate by reporting in favor of urban

consumers? If so, what are the possible factors which cause this media bias?

The main objectives of this study are to illustrate the perception of the ‘food crisis’

concept on the media, to explore the differences in media attention between consumers and

farmers, to explore possible factors that lead to media bias, and to make a dynamic and

progressive contribution to further applied economics researches.

5

2. A Framework for DiscussionThe basic principles of agricultural price changes and their effects on consumers and

producers are shortly summarized in The Right Price of Food by Swinnen (2010). As a starting

point, Swinnen’s work is followed to set a framework on high versus low food prices. A simple

model of an open economy with two groups is considered: producers and consumers of food,

where prices are determined at the world market, and local production or consumption do not

have any impact on global prices (i.e. small country assumption). The basic framework to assess

the effects of changes in prices for staples is straightforward. Consumers will lose from price

increases, producers will gain, and vice versa when price decrease. As many households in

developing countries will be both producers and consumers, the net impact effect of price

changes will be determined by which effect is greater: whether the household is a net consumer

or a net producer (Deaton, 1989). The effect of world market price changes may differ for

countries – due to different trade policies, institutions, and the industrial organization of the food

chain – or local production and consumption may affect local prices, the exogenous shocks may

also be caused by nature or by people, or short-run effects may differ from the long-run effects.

All these extensions do not fundamentally change the basic principle of the simple model: when

prices go up, consumers lose and producers gain, and vice versa. Hence, when rich countries

increase (reduce) export subsidies which leads to a decline (increase) in world markets, this will

benefit (hurt) urban consumers and net -consuming rural households in poor countries, and hurt

(benefit) net-producing rural households in poor countries (Swinnen, 2010: 4-5).

However, a straightforward implication of these basic principles is that consumers lose

and producers gain in post-2007 period, and vice versa in pre-2007 period; we do not find them

reflected in most media discourses. There has been hardly any mention of the losses of rural

farmers from low food prices, but a very strong emphasis on food price inflation and its poverty

impacts on consumers, especially on urban populations. Additionally, there has been very little

emphasis on farmers even when they were supposed to gain from high food prices.

3. Mass Media Communications

3.1. Media: A window or a mirror of public opinion?There are two simple and contrasting approaches to the relationship between mass media

content and public opinion. The first approach regards media as a shaper, arguing that the media

6

affects and influences public opinion. The other approach regards media as a mirror of public

opinion, arguing that the media reflects society values, identities, social patterns, events or

behaviors which are already identified by society (Hodkinson, 2011). In this logic, public

officials react to media news because they see it as a reflection of public opinion (Kim, 2005: 3).

These approaches which argue that media content either influences or mirrors public opinion are

both simplistic. It is believed that ‘a more useful starting point is to conceive it as an ongoing

process whereby selective media representations constantly feed into and are themselves fed by

the make-up and character of the public’ (Hodkinson, 2011: 6). This paper uses a mixture of

approaches which corresponds to an interactive circle of public and media.

3.2. Debates on Food Prices: Media DiscoursesAs previously stated, that media both mirrors and influences public opinion, taking a look

at the statements from the international newspapers in pre- and post-food crisis periods is a good

start to make latter analyses clearer.

The Guardian

“There is no world food crisis, as Kofi Annan, secretary-general of the UN, repeated recently.

Instead, there is an enduring reluctance, demonstrated once again in the feeble but disputed Franz

Fischler plan to reform the CAP, to cut the agricultural subsidies and protectionism that bankrupt

the farmers of the developing world. Eight hundred million people still go hungry because they

cannot afford to grow or buy their own food.” (2002)

“Crisis talks on global food prices… World leaders are to meet next week for urgent talks aimed

at preventing tens of millions of the world's poor dying of hunger as a result of soaring food

prices…The urgency of the meeting follows historic spikes in the price of some staple

foods…The world's urban poor have been hit hardest, sending a wave of unrest and instability

around the world. Thirty-seven countries have been hit by food riots so far this year.” (2008a)

“Europe and the global food crisis…EU food production levels must be held steady – for the sake

of Europeans and of people in the world's poorest countries. The world has been shaken by

unprecedented spikes in food prices, by hunger riots, and by social tensions that demonstrate that

food supplies have returned as a source of insecurity…By 2050, it is estimated that there will be

7

9 billion people on earth, so the need for food may double – primarily among urban populations

in the world's poorest countries.” (2008b)

The Economist

“Six million Ethiopians are going hungry. In all, 5.9m of Ethiopia's 65m people need food aid or

other assistance, according to the World Food Programme.” (2002a)

“The director for East and Southern Africa for the UN World Food Programme confirmed that at

least half of the 12m people in Zimbabwe will need some kind of food assistance, and that

malnutrition was increasing.” (2002b)

“Ethiopia's prime minister, Meles Zenawi, said that his country faces a famine even more

destructive that the disaster of 1984 which so appalled western television viewers. Some 15m

Ethiopians, he said, will need food aid by early next year, out of a population of about 65m.”

(2002c)

“After the year of food crisis, the year of the farmer… In 2008, the World Bank reckoned that

higher food prices drove 100m people into poverty. That may have been a bit of an exaggeration

because many rural parts of developing countries also benefited from dearer food. But the urban

poor suffered a great deal and the majority of developing countries were hit by higher inflation

and lower living standards.” (2008)

The Mail & Guardian

“Nearly half the population of Zimbabwe is facing hunger and needs food assistance as the

country’s food emergency deepens, a famine early-warning group reported…(M)alnutrition and

related diseases are expected to peak from January to March, ahead of the next harvests. In many

urban and rural areas, families are forced to drastically reduce food consumption “or spend the

whole day without having a meal at all”, while programmes to help the elderly, the chronically

ill, orphans and other vulnerable groups are grossly inadequate.” (2005a)

“Food security is a continuing problem in Malawi where an estimated 60% of rural households

are unable to meet nutritional needs. More than 60% of Malawians live on less than one US

dollar a day” (2005b)

8

“We consider that the dramatic escalation in food prices worldwide has evolved into an

unprecedented challenge of global proportions,” the United Nations said in a statement. This had

become a crisis for the world’s most vulnerable people, including the urban poor, it said after a

meeting of 27 international agency heads in the Swiss capital, Berne, to chart a solution to food

price rises that have caused hunger, riots and hoarding in poor countries…‘‘We are urging

countries not to use export bans. These controls encourage hoarding, drive up prices and hurt the

poorest people around the world who are struggling to feed themselves,” World Bank (Zoellick)

said. Rising food prices have hurt the growing mass of urban poor in developing countries most,

but have done little to help poor farmers, who are wary of growing more food because of the

concurrent rise in fuel and fertilizer costs, Zoellick said.” (2008)

As can be seen above, the newspapers that publish articles on food policy issues have

started to create a new story line on the food crisis debate. In the post-2007 period, high food

prices were argued to cause poverty, hunger and food riots among urban populations.

Henceforward, ‘food crisis’ has become the new story line on the current debate. On the contrary,

poverty and hunger were affecting rural populations badly in the pre-2007 period, which was not

then named as a crisis.

One can argue that statements are taken out of context so that the assertions do not reflect

the true key messages in the entire reports. It is of course better to read full documents to provide

deeper information. Quoting is still convenient and a good start for media analysis since

developing key messages is crucial for the media and the rest of the media reports do not deviate

from the key messages, but confirm them. However media mention poverty and hunger

discourses in both periods; a normal situation became a ‘crisis’ after 2007 because the target

group of poverty and hunger shifted from rural to urban populations.

4. MethodologyAs stated previously, these quotes were used as a starter to discuss the issue, and they

provide a better understanding for latter analysis. There is no clear consensus as to what

discourses are or how to analyze them. Different perspectives offer their own suggestions and, to

some extent, compete to appropriate the terms ‘discourse’ and ‘discourse analysis’ for their own

definitions (Jørgensen & Phillips, 2002:12). According to experts on DA, texts are not

9

individually meaningful (Hardy & Phillips, 2002; Neuendorf, 2004: 33). This notion strikes at

the heart of a primary commonality between DA and CA. Both are concerned with drawing

conclusions about some aspect of human communication from a carefully selected set of

messages. How they do so is rather different, but ultimately their findings can fit together quite

nicely, providing a good example of triangulation of methods a highly desirable situation.

(Neuendorf, 2004: 33). Since there is no clear consensus as to how to analyze texts and the paper

examines a huge number of articles over the 13-year period, CA, which gives a chance to use

statistical methods and go one step further by transforming texts into numbers – in terms of

frequencies of the statements by time, is chosen as a complement to DA in the present study. It

then uses the data derived from CA for further statistical analysis.

4.1. Data SourceThe paper anayzes the articles reported on the food crisis. These articles were published in

three international newspapers, since January 2000 till mid-April 2013; more specifically, before,

during and after the ‘global food crisis’. In this paper, 78, 159 and 100 reports, respectively taken

from the online versions of The Economist, The Guardian and The Mail & Guardian, will be

investigated as representatives of the entire mass media. However, the paper analyzes a huge

number of online media articles; there can still be a large number of them in this regard to select

from. Due to time restrictions a set of criteria is adopted to set the limits of the most related

corpus to be examined. Text selection was based on standard techniques, such as keyword

computer searches: for instance, food crisis, world food crisis, global food crisis. The reason for

choosing these three newspapers as representatives of the entire group of newspaper magazines

are their being highly ranked and creditable all over the world; having printed and online formats

that give audiences the opportunity to search online; and being owned by companies, and not by

goverments which shows that they represent the views of their stakeholders and the audiences. At

this point, one can argue that The Mail & Guardian does not represent the international media.

However, we would argue that it does represent African media and provides a more

comprehensive approach for the research, since African countries are believed to be the most

vulnerable to volatile food prices.

Up to this point, the data sources have been circumscribed. It is now important to decide

how such data in text form is going to be examined in order to answer the research questions.

10

4.2. Creating Data: Coding the Media FramesIn this paper ATLAS.ti, a qualitative data analysis software, is used to code the media

frames as keywords. Although coding is a simple method to follow, is a crucial step to start.

What was done was basically to read the reports and to code the paragraphs or sentences as

keywords which are important for the analysis, such as food price inflation, global/world food

crisis, poverty, hunger, malnutrition, famine, food insecurity, farmer, etc. For instance, a text like

‘nearly half the population of Zimbabwe is facing hunger…’ is coded as ‘hunger’. Coding

process was repeated several times for each code in order to minimize the subjectiveness of the

coder. When the coding is done, the next step is to put the total amounts of each code in order by

the publication date of the reports that the codes were stated. The number of occurrences of the

codes (i.e. frequencies) bring into being the (numerical) data for the statistical analysis.

4.3. Model and HyphothesisThe analysis contiues with the Ordinary least-squares (OLS) regression model which has

‘the code of food crisis’ (i.e. annual frequencies of food crisis statements) as a dependent

variable. The independent variables are the annual frequencies of codes which represent food

price inflation, food riot, poverty, hunger, malnutrition, famine, farmer(negative),

farmer(positive), africa food crisis, a dummy variable (year 2008), and time trend.

The regression equations for the three newspapers are:

= + +⋯+ u (1)

= + +⋯+ u (2)

= + +⋯+ u (3)

where Y, and u are T-vectors, a, b, …, n are the T*k matrix of regressors, and x is the k-vector of

the parameters. OLS minimizes the sum of the squared residuals.

It is hypothesized that the media will pay more attention to the consumers than to the

producers when food prices change, and if this hypothesis is accepted, the results will be

significant for the concepts which are related with consumers, such as poverty, hunger, food riots,

etc., and they will have a positive relationship with food crisis; while the results will not be

11

significant for the concepts which are related with farmers, such as farmer(negative), and

farmer(positive).

5. Key ResultsIn this paper, the articles which were reported on the food crisis in three international

newspapers, since January 2000 till mid-April 2013, more specifically before, during and after

the ‘global food crisis,’ are examined. 78, 159 and 100 reports, respectively taken from the online

versions of The Economist, The Guardian and The Mail & Guardian are investigated as

representatives of the entire mass media.

5.1. Results from content analysis

Figure 1. CA for the frequency of the code ‘global food crisis’ stated in The Guardian (blue), The Mail &

Guardian (grey) and The Economist (black) by year.

As shown in Figure 1, the global world crisis has been argued in the media since 2007,

having its peak in 2008. The numbers of articles which were published in the pre-2007 period and

also the frequencies of the code of “global food crisis” in these articles are too small to be

ignored. In the pre-2007 period, the newspapers mentioned ‘food crisis’ only 9 times while they

mentioned it 384 times in the post-period, 221 among these were stated in 2008. In post-2007,

-20

0

20

40

60

80

100

120

1998 2000 2002 2004 2006 2008 2010 2012 2014

12

especially in 2008, all the newspapers suddenly started to talk about a ‘food crisis’ and during all

the post- period, they kept talking about a ‘food crisis’.

Figure 2. CA for the frequency of the codes ‘global food crisis’ (black) and ‘food price inflation’ (grey)

by year.

What was reasoning this sudden change in communication? The analysis shows that it

was ‘food inflation’ what leads to food crisis talks (Figure 2). There is a dramatic change in the

emphasis on ‘food price inflation’ in the pre- and after- periods. In the post-2007 period, the

newspapers mentioned ‘food price inflation’ 520 times while they mentioned it only 5 times in

the pre-period. More than half of the statements in post- period occurred in 2008. The results

show the change in emphasis on ‘food price inflation’ is more dramatic than the change in

emphasis on ‘food crisis’ for pre- and post- periods. Also, changes in emphasis on both codes

move in the same direction. For example, in 2010, a second increase in food prices after a shock

increase in 2008 made the media talk about a ‘food crisis’ again. We should also take into

consideration that in some articles, ‘food price inflation’ was used as a complement to or even

instead of ‘food crisis’. ‘Food crisis’ has been mentioned in most of the articles where ‘food price

inflation’ has been mentioned. Up to this point, the situation is analyzed only from the

consumer’s side. The media (the public) name the situation of the changes in prices, which starts

to hurt consumers, as a ‘crisis’.

-50

0

50

100

150

200

250

300

350

1998 2000 2002 2004 2006 2008 2010 2012 2014

13

Figure 3. CA for the frequency of the codes ‘global food crisis’ (grey), ‘farmers-positive’ (black) and

‘farmers-negative’ (blue) by year.

Positive farmer-side discourses, such as:

“…Farmers in developed countries are the winners of food crisis…,

…After the year of drought, the year of farmers…

…Farmers benefit from high food prices…”

refer to ‘farmers-positive’ while negative farmer-side discourses, such as:

“…Farmers who were already poor became poorer…,

…Poor farmers could not benefit from the high food prices…,

…Farmers in developing countries continuing suffering in the year of farmer…”

refer to ‘farmers-negative’.

As said before, consumers will lose from price increases; producers will gain, and vice

versa when price decrease. Since the food prices were very low in the pre-2007 period, one

would expect the media to mention the losses of farmers; however, we do not find this

straightforward principle’s implication reflected in media arguments. In the pre- period, farmers

were mentioned only 12 times which consist of 11 negative discourses and 1 positive discourse.

We daresay that there was a lack of media coverage of the losses of the farmers. Additionally,

one would expect the media to mention the benefits of farmers in the post-period, when the food

prices increased dramatically. However positive discourses start to appear, negative discourses

still exist, to an even larger extent. Surprisingly, negative discourses concerning farmers (99

statements) are more than positive discourses (44 statements) in the post- period. Negative

-50

0

50

100

150

200

250

1998 2000 2002 2004 2006 2008 2010 2012 2014

14

discourses refer to farmers in developing countries, while positive discourses refer to farmers in

developed countries. Considering that some of the farmers in developing countries are net buyers

of food and most of the farmers hurt by increasing energy prices, it is not surprising anymore that

negative media coverage of farmers in the post- period is quite large. However, the emphasis on

farmers in the post- period is much stronger than the pre- period; the overall emphasis on farmers

is very little comparing to the overall emphasis on the concepts concerning consumers. In total,

farmers were mentioned 155 times in over 13 year period. In other words media coverage of

consumers is 6 to 8 times (depends on taking ‘‘food price inflation’’ into account) bigger than

media coverage of farmers.

Table-1. Elements of media communication before and after food crisis

Table 1 demonstrates the key discourses which are important in the food crisis debate and

summarizes some of the key findings prior to and after the food crisis. In 2008, in media

discourses, the strongest emphases were on food price inflation, food crisis, poverty and hunger,

which were practically not mentioned in the pre-2007 period. As previously discussed, ‘food

STORY LINES Note: 0: not mentioned, 1-10: from the weakest to the strongest emphasis

DISCOURSES

africa's food crisis 1 1 2 1 1 1 1 1 2 1 1 1 2 1

biofuels 0 0 0 0 0 0 0 2 6 1 1 1 1 1

famine 1 1 2 1 1 1 0 1 2 2 1 2 2 0

farmers(negative) 0 1 1 0 1 1 1 1 4 1 1 1 1 1

farmers(positive) 0 0 0 0 1 0 0 0 2 1 1 1 1 0

food aid 1 1 2 1 1 1 1 1 3 1 1 1 1 1

food crisis 0 1 0 0 1 1 0 1 9 3 3 2 3 1

food price inflation 0 1 1 0 1 2 0 2 10 2 5 2 4 1

food riots 0 1 1 0 0 0 0 1 4 1 2 1 1 1

food security 0 1 1 0 1 1 0 1 3 2 2 2 1 1

hunger 1 1 1 0 0 1 0 1 7 1 2 2 2 1

malnutrition 0 0 1 0 0 1 0 1 2 0 1 1 1 1

poverty 0 0 1 0 1 1 0 1 9 1 2 1 1 1

time (t): 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

15

crisis’ became a new storyline in the recent debate (Figure 1); the talks on food crisis were

referring mostly to a local food crisis due to bad weather conditions and bad politics, mainly in

African countries (Table 1). It is also remarked that ‘food price inflation’, having the strongest

emphases in all in 2008, led to ‘food crisis’ talks (Figure 2 and Table 1). Increasing food prices

helped the media to attach attention to poverty and hunger (Table 1). The poor who spend most

of their income on food became poorer and those in the cities started to protest against high food

prices. So, food riots became an important issue in the post-2007 period (Table 1). In mid-2008,

The Guardian reported that 37 countries were hit by food riots (2008a). Throughout the post-

period, more than 40 countries experienced food riots, mostly in the urban areas.

There is a vast amount of literature on the causes of the food crisis. One of the most

important factors which caused increases in agricultural commodity prices was biofuel

production (Abbott et al., 2008; Chand, 2008; Mitchell, 2008; Schnepf, 2008; USDA 2008; von

Braun, et al., 2008;). Since 2006, biofuel production has surged with the rapidly rising energy

prices and improved bioenergy conversion technologies. Energy and agricultural markets became

closely linked (Abbott et al., 2008; Chand, 2008; Schmidhuber, 2006), and usage of crops for

ethanol and biodiesel became large enough to influence world prices. There is a dramatic change

in the emphasis on biofuels in pre- and post-periods (Table 1).

Some other discourses which are stated in Table 1, such as famines, food aid,

malnutrition and Africa food crisis were not as ‘famous’ as the others in either period. Africa has

always been suffering from hunger, poverty and malnutrition, and is often hit by many local food

crises and famines.

5.2. Results from the OLS regression

As stated previously, it is hypothesized that the media will pay more attention to

consumers than to producers when food prices change; if this hypothesis is accepted, the results

will be significant for the concepts which are related with consumers, such as poverty, hunger,

food riots, etc., and they will have a positive relationship with food crisis. Each explanatory

variable in the model is assessed: Coefficient, Probability or Robust Standard Errors (HAC), and

Variance Inflation Factor (VIF). The T-test is used to assess whether or not an explanatory

16

variable is statistically significant. Each consumer-side explanatory variable is expected to be

significant and the signs of the coefficients to be positive, to have a positive relationship with the

dependent variable. In contrast, farmer-side explanatory variables are expected to be non-

significant. The models do not include first order autocorrelation (i.e. the Durbin-Watson test

statistics are very close to 2 in all models). The models for The Economist and The Mail &

Guardian do not include collinearity problem (i.e. the variables which have VIF values bigger

than 10.0 are omitted), while the other model includes a negligible collinearity problem (VIF

value for food price inflation is 12.699). According to R-squared, in all models, all the

explanatory variables modeled using regression explain more than 85% of the variation in the

dependent variable (food crisis).

Table 2. Summarized OLS Results *

Independentvariables/Newspapers

The Guardian The Economist The Mail & GuardianSign of

coefficientLevel of

significanceSign of

coefficientLevel of

significanceSign of

coefficientLevel of

significancefood price inflation + ns + ns + ***

food riot + *** + ns - **poverty + ns + * + nshunger + ** + ** + *

malnutrition - ns + ns - **famine + * + ns - ns

farmer(negative) - ns + ns - nsfarmer(positive) - ns + ns + ns

time trend + ** + *** - nsdummy 2008 + ** + ns + ns

africa food crisis - ns - *** - ns

R-squared 0.865041 0.851133 0.873680*ns: nonsignificant, *: 0.1< p value<0.05, **: 0.01<p value<0.05, ***: p value<0.01

As shown in Table 2, the code of ‘hunger’ is significant with a positive coefficient for all

the newspapers analyzed. The more we talk about hunger, the more we talk about food crisis. The

other codes concerning consumers, such as ‘food price inflation’, ‘food riots’, ‘poverty’,

17

‘malnutrition’, and ‘famine’ are all significant for at least one of the newspapers and have

positive coefficients, except malnutrition. The codes concerning farmers are not significant at all.

The emphasis on ‘food crisis’ do change with respect to time. In other words, time is an

important determinant and positively affects the dependent variable. We can also say that the

emphasis on ‘food crisis’ increased in 2008. African food crises are believed to be local or

regional problems mostly caused by bad weather conditions and bad policies. The result from the

analysis of The Economist show that the media attention shifted from African food crises to

global food crisis by time. However it is not clearly shown by the results, The South African

newspaper (The Mail & Guardian) differs in coverage of African problems and it kept reporting

on ‘African food crises’ during the global food crisis.

Finally, the regression results substantiate the results from DA and CA by supporting

research hypotheses. In contrast with consumers, farmers do not play a significant role in the food

crisis debate. Food crisis talks become popular when consumers are hurt by food price changes.

6. DiscussionAs a consequence of increasing food prices, the poor who spend most of their income on food

became poorer and the ones in the cities started to protest against high food prices. Worldwide

food riots became an important issue in the post-2007 period. Hence, the media started to pay

more attention to the consequences of high food prices on consumers, such as poverty and hunger

(Table 1). Eventually ‘food crisis’ discourse which was led by ‘food inflation’ discourse became

a new storyline in the post-2007 period (Figure 2 and Table 1). Even though there were poverty

and hunger problems in the pre-2007 period and many of the rural farmers were suffering from

low food prices, there was an absence of media coverage for the food crisis (Figure 3 and Table

1) – with some exceptions in the context of an African food crisis (Table 1). However, those

discourses were limited to a country or a region, and the causes were bad weather conditions and

bad governance. After 2007, we started to hear about a ‘global food crisis’ which affects all the

poor badly, as well as the African poor (Figure 1), and we also kept hearing about an ‘African

food crisis’.

Media discourses are biased by emphasizing the negative welfare effects of high food prices

on consumers, especially in urban areas and ignoring the negative welfare effects of low food

18

prices on farmers (Figure 3). Selective media coverage contributes to an irrational allocation of

short-term emergency relief because coverage is determined by factors other than humanitarian

need (Jakobson, 2000).

The media are also selective in reporting farmers’ benefits from high food prices. Since

consumers are losing because of high food prices, one would expect more media attention to the

benefits of farmers. One can argue that farmers in developing countries could not benefit much

from high prices. We would still argue that the media do not pay that much attention to farmers,

regardless of how much and in which direction they are affected by food price changes (Figure 3

and Table 1). But they do pay attention to consumers when price changes affect them (Table 1).

In contrast with consumers, farmers do not play a significant role in the food crisis debate.

Many factors might lead to this media bias. First, a disproportionate amount of attention

which has been paid to the problems of urban consumers can be explained by urban bias.

According to (FAO, 2008 :4), ‘Riots and civil disturbances, which have taken place in many low-

and middle-income developing countries, signal the desperation caused by soaring food and fuel

prices for millions of poor and also middle-class households.’ The groups who usually respond

politically with strikes, protests or riots to the negative income effects of food price changes are

urban consumers, not rural farmers; it is easier to mobilize the urban populations who are already

concentrated in the cities. However, Hendrix et al. (2009) argues that farmers are acutely aware

of the political gains that come from concentrating collective action in cities; we agree with von

Grebmer et al. (2008) concerning the farmers, who usually live in distant rural areas, suffer

silently for a while. A lack of protests may not correctly depict the severity of impact on the

poorest of the poor (von Grebmer et al., 2008). It may be that a similar urban bias effect plays a

role in drawing policy attention through global media (Swinnen, 2010), since dissatisfied urban

consumers have an important role in driving government policy on the issue. Policymakers favor

programmes that protect city dwellers from feeling the full brunt of food price increases in large

part because these have a pacifying effect on this urban discontent (Cohen and Garrett, 2009).

Second, media attention is typically concentrated around “events”. Events trigger media

attention by a combination of dynamic demand and supply effects. The importance of events is

driven on the demand side by the increased interests of consumers, in some cases through their

anticipation of important implications, and on the supply side by the availability or lower cost of

19

stories, including pictures and illustrative material. Some of the forces that cause increased media

attention on an issue at a certain moment also contribute to reduced attention to other moments,

as other activities or events take over (Swinnen and Francken, 2006). After a period of decreasing

food prices for almost 50 of the last 60 years, and low food prices during the last 20 years, food

prices started to increase dramatically. Only three price spikes have been seen in the last 60 years,

together lasting just about ten years (Aksoy & Beghin, 2004; Aksoy & Hoekman, 2010: 2). This

breaking news has been widely reported.

It is believed that media consumers tend to be more interested in negative coverage, and

they choose to consume more negative stories than positive stories (the so-called “bad news

hypothesis’’). McCluskey and Swinnen (2010) state: “Consumers use the information from

positive media stories to take advantage of opportunities from positive shocks and use the

information from negative stories to avoid negative shocks. If utility is concave, the marginal loss

in utility from not consuming the first bad news story is greater than the marginal gain in utility

from consuming the first positive news story. As a result, consumers will choose to consume more

negative stories than positive stories.” This demand effect of the media market drives more

negative coverage in reporting.

The importance of attracting large numbers of readers may in itself also lead to bias

(McCluskey and Swinnen, 2010). Strömberg (2004) argues that increasing returns to scale in

media markets induce the media to cover the issues that are of interest to larger groups.

A combination of all these factors creates a bias in reporting. Mass media news bias will

translate into a bias in public policy. Small groups will receive less favorable policies because of

the provision of information by mass media firms (Strömberg, 2004: 281).

Swinnen (2010) and Swinnen et al. (2011) refer to bias in food policy communication of

development organizations and NGOs, such as FAO, IMF, WB, OECD, IFPRI, Oxfam and

Action Aid, which are supposed to aim to increase welfare and reduce poverty around the world.

They argue that these organizations do not recommend coherent policy messages because they

are biased by the communications of potential donors in fund-raising. Their policy messages are

not correct when prices go up and down because they only report the negative effects and ignore

the positive effects of price shocks. For instance, the organizations claim that low food prices

caused poverty and food insecurity prior to the food crisis, and high food prices caused poverty

and food insecurity after the food crisis. When conditions change, the focus group which is

20

affected by the new conditions change, so that policy recommendations change. They also argue

that organizations attract media coverage by reporting negative welfare implications in their

policy communications and media induce them to act to affect the food policy. The present paper

and above-mentioned papers analyze changing discourses prior to and after the food crisis, and

argue that there are biased actors which can lead to a bias in food policy. These papers use

discourse methods based on textual analysis while the present paper uses statistical techniques

combined with discourse methods.

7. Concluding RemarksIn the light of all the above-mentioned facts, there are several reasons which may explain

why media discourses are biased by emphasizing the negative welfare effects of high food prices

on consumers, especially in urban areas, ignoring the negative welfare effects of low food prices

on farmers. Yet another very important question based on the research findings is how big the

effect of media bias on policy bias is. Swinnen (2010) argues that it is difficult to answer this

question since it depends on various assumptions regarding the processing of these sets of

information by voters, policy-makers and the organizations themselves, the type of welfare

function one has in mind, and the political economy of policy decisions at various levels

(Swinnen 2010: 26).

Table 3. The Political Economy of Food: Players, Fears, Hopes and Options

Actors Fears Hopes Options

Farmers Bad weatherCrop failureLow prices

High yieldsGood prices

Jobs off-farm

Switch cropsNew technology“Demonstrate”

Consumers Food shortagesPrice spikesLoss of jobs

Safe food in marketsStable prices (safety nets)

HoardingDiversify diet

RiotsMedia Loss of market power

No one listensBad news

Stakeholders’ satisfactionAudiences’ satisfaction

Profits

Create moral panicSwitch agenda

Lobbying

Politicians Food riotsNot re-elected

Coup/forced exit

Low pricesFarmers’ votes

Consumers’ votes

Tariffs and price stabilizationSubsidies

Safety nets[Based on tables in The Economist (2011), and Timmer (2011): 26]

21

In Table 3, the main actors and their fears, hopes and possible actions, which have mostly

been mentioned in this study to keep a balance of forces on the food policy debate, are given.

Any changes in food policy will depend on these four players’ actions and the balance of forces.

As previously discussed, media bias may lead to policy bias, since politicians see biased media as

a mirror of public opinion. A possible policy bias may cause an unequal welfare distribution

among societies and misdirect development policies.

This study sheds light on media bias in the food crisis debate and possible factors which

cause their shifted discourses between the pre-2007 and post-2007 periods. It will hopefully lead

to further studies to explore the effects of shifted discourses on policy making, and hence, on

welfare and the future development process of nations.

22

References

Abbott, P., Hurt, C. & Tyner, W. (2008). ‘What’s driving food prices?’ Issue Report. Farm Foundation.

Ahmed, A. U., Vargas Hill, R., Smith, L. C., Wiesmann, D., Frankenberger, T. (2007). The World’s Most Deprived:

Characteristics and Causes of Extreme Poverty and Hunger. 2020 Discussion Paper 43. International Food

Policy Research Institute (IFPRI), Washington DC.

Akaike, H. (1974). A new look at the statistical model identification, IEEE Transactions on Automatic Control AC-

19: 716–723.

Aksoy, M. A. & Hoekman, B. (2010). Food Prices and Poverty: Introduction and Overview. In: Aksoy, M. A. &

Hoekman, B. (eds), Centre for Economic Policy Research. Food prices and rural poverty. CEPR :1-24

Aksoy, M. A., & Isik-Dikmelik, A. (2010). Are Low Food Prices Pro-Poor? Net Food Buyers and Sellers in Low-

Income Countries. Centre for Economic Policy Research, 113.

Aksoy, M.A., and J. C. Beghin (eds.), 2004. Global Agricultural Trade and Developing Countries. Washington, DC:

World Bank

Berelson, B. (1952). Content analysis in communication research. New York, NY, US: Free Press.

Chand, R. (2008). The Global Food Crisis: Causes, Severity and Outlook. Review of Agriculture, Economic &

Political Weekly, June 28, 2008: 115-123.

Cohen, M. & Garrett, J. (2009). The Food Price Crisis and Urban Food (In)security, Human Settlements Working

Paper Series, London, International Institute for Environment and Development.

Deaton, A. (1989). Rice Prices and income Distribution in Thailand: A Non-parametric Analysis. The Economic

Journal, 99(395), 1-37.

Dessus, S., Herrera, S., & De Hoyos, R. (2008). The impact of food inflation on urban poverty and its monetary cost:

some back‐of‐the‐envelope calculations. Agricultural Economics, 39(s1), 417-429.

FAO (2008). The State of Food Insecurity in the World 2008. FAO, Rome.

FAO (2012). High and volatile food prices: Supporting farmers and consumers. CAADP Policy Brief 08, Future

Agricultures Consortium.

Hardy, C. & Phillips, N. (2002). Discourse analysis: Investigating Processes Of Social Construction. Thousand

Oaks, CA: Sage Publications.

Hardy, C., Harley, B., & Phillips, N. (2004). Discourse analysis and content analysis: Two solitudes. Qualitative

Methods, 2(1), 19-22.

Headey, D., & Fan, S. (2008). Anatomy of a crisis: the causes and consequences of surging food prices. Agricultural

Economics, 39(s1), 375-391.

Hendrix, C., Haggard, S. and Magaloni, B. (2009). Grievance and opportunity: Food Prices, Political Regime, and

Protest. Paper presented at the International Studies Association convention, New York, 15-18 February.

Hodkinson, P. (2011). Media, Culture and Society: An Introduction. Sage Publication.

IMF (2008). Food and fuel prices: Recent developments, macroeconomic impact, and policy responses. Fiscal

Affairs, Policy Development and Review, and Research Departments, International Monetary Fund,

23

Washington DC.

Ivanic, M., & Martin, W. (2008). Implications of higher global food prices for poverty in low‐income

countries. Agricultural Economics, 39(s1), 405-416.

Jakobson, P. V. (2000). Focus on the CNN effect misses the point: the real media impact on conflict management is

invisible and indirect. Journal of Peace Research, 37(2), 131-143.

Jørgensen, M. W., & Phillips, L. J. (2002). Discourse analysis as theory and method. Sage.

Kim, J. S. (2005). Media coverage and foreign assistance: The effects of U.S. media coverage on the distribution of

U.S. official development assistance (ODA) to recipient countries. Washington, DC: Georgetown Public

Policy Institute.

McCluskey, J.J. & Swinnen, J. F. M. (2010). Media economics and the political economy of information. In: Coen,

D., Grant W. & Wilson G. (eds), The Oxford Handbook of Government and Business. Oxford: Oxford

University Press.

Mitchell, D. (2008). A Note on Rising Food Prices. Policy Research Working Paper 4682, The World Bank

Development Prospects Group, July 2008.

Neuendorf, K. A. (2004). Content analysis: A contrast and complement to discourse analysis. Qualitative methods:

Newsletter of the American Political Science Association Organized Section on Qualitative Methods, 2(1), 33-

36.

Schmidhuber, J. (2006). Impact of an increased biomass use on agricultural markets, prices and food security: A

longer-term perspective. International symposium of Notre Europe, Paris.

Schnepf, Randy (2008). High Agricultural Commodity Prices: What Are the Issues? Congressional Research Service

(CRS) Report for Congress, May 6, 2008.

Strömberg, D. (2004) ‘‘Mass media competition, political competition, and public policy,’’ Review of Economic

Studies 71: 265-84.

Swinnen, J. (2010). The right price of food: Reflections on the political economy of policy analysis and

communication (No. 259). LICOS Discussion Paper.

Swinnen, J. F., & Francken, N. (2006). Summits, Riots and Media Attention: The Political Economy of Information

on Trade and Globalisation. The World Economy, 29(5), 637-654.

Swinnen, J. F., McCluskey, J., & Francken, N. (2005). Food safety, the media, and the information market.

Agricultural Economics, 32(s1), 175-188.

Swinnen, J. F., Squicciarini, P., & Vandemoortele, T. (2011). The food crisis, mass media and the political economy

of policy analysis and communication. European Review of Agricultural Economics, 38(3), 409-426.

The Economist (2002a). Famine in Ethiopia: Drought, death and taxes (2002, May 9).

http://www.economist.com/node/1316910

--- (2002b). Mugabe’s famine (2002, August 14). http://www.economist.com/node/1280403

--- (2002c). Bad weather, bad government (2002, November 11) http://www.economist.com/node/1440625

--- (2008). Old Macdonald gets some cash: After the year of food crisis, the year of the farmer (2008, November

19). http://www.economist.com/node/12494668

24

--- (2011). Hungry for votes: How much do rich governments really worry about feeding the world? (2011,

January 29).

The Guardian (2002). Fat is a fatuous issue (2002, May 30).

http://www.guardian.co.uk/lifeandstyle/2002/jun/30/foodanddrink.maryriddell?INTCMP=SRCH

---(2008a). Crisis talks on global food prices (2008, May 27).

http://www.guardian.co.uk/environment/2008/may/27/food.internationalaidanddevelopment?INTCMP=SRCH

---(2008b). Europe and the global food crisis (2008, November 25).

http://www.guardian.co.uk/commentisfree/2008/nov/25/eu-food?INTCMP=SRCH

The Mail & Guardian (2005a). Zimbabwe: 5,8m go hungry as food crisis worsens (2005, January 28).

http://mg.co.za/article/2005-01-28-zimbabwe-58m-go-hungry-as-food-crisis-worsens

--- (2005b). Malawi asks for aid to avert food crisis (2005, May 10).

http://mg.co.za/article/2005-05-10-malawi-asks-for-aid-to-avert-food-crisis

--- (2008) UN and World Bank to tackle food crisis (2008, April 29).

http://mg.co.za/article/2008-04-29-un-and-world-bank-to-tackle-food-crisis

Timmer, P. (2011). Managing price volatility: approaches at the global, national, and household levels. In Center on

Food Security and the Environment. Stanford Symposium Series on Global Food Policy and Food Security in

the 21st Century (FSE Stanford), Stanford University, Stanford, CA.

http://iisdb.stanford.edu/evnts/6553/Timmer_final_11.pdf

U. S. Department of Agriculture (USDA), 2008. “Agricultural Projections to 2017,” February 2008.

http://www.ers.usda.gov/media/274754/oce20081_1_.pdf

von Braun, J. (2008). Rising Food Prices: What Should Be Done? IFPRI Policy Brief, The International Food Policy

Research Institute (IFPRI).

von Grebmer, K., Fritschel, H., Nestorova, B., Olofinbiyi, T., Pandya-Lorch, R., Yohannes, Y. (2008) ‘Global

Hunger Index: The challenge of hunger 2008’ Deutsche Welthungerhilfe, International Food Policy Research

Institute and Concern Worldwide, Bonn, Washington, DC, and Dublin.

Wodon, Q., & Zaman, H. (2008). Rising food prices in sub-Saharan Africa: Poverty impact and policy responses.

World Bank Policy Research Working Paper 4738, October.

Wodon, Q., C., Tsimpo, P. , Backiny-Yetna, G. , Joseph, F. Adoho and Coulombe, H. (2008). Potential impact of

higher food prices on poverty : summary estimates for a dozen west and central African countries. Policy

Research Working Paper Series 4745, The World Bank, Washington DC.

World Bank, (2008). Addressing the Food Crisis: The Need for Rapid and Coordinated Action. Background paper

for the Group of Eight Meeting of Finance Ministers, Osaka, June 13-14, 2008. The World Bank, Washington

DC. http://www.worldbank.org/html/extdr/foodprices/pdf/G8_food%20price%20paper.pdf.

Zezza, A., Davis, B., Azzarri, C., Covarrubias, K., Tasciotti, L., & Anriquez, G. (2008). The impact of rising food

prices on the poor. Unpublished manuscript. Food and Agriculture Organization, Rome.

Zito, G. V. (1975). Methodology and meanings: Varieties of sociological inquiry. New York: Praeger.

25

Annex 1. Ordinary Least Squares (OLS) Results*

Model 1 - The Guardian: OLS, using observations 2000:01-2013:04 (T = 160)Dependent variable: foodcrisis

HAC standard errors, bandwidth 4 (Bartlett kernel)Coefficient Std. Error t-ratio p-value

const -0.0501739 0.137697 -0.3644 0.71609time 0.00441829 0.00193061 2.2885 0.02352 **africafoodcrisis -0.292311 0.254728 -1.1475 0.25301famine 0.253832 0.148635 1.7078 0.08978 *farmersnegative -0.340239 0.492476 -0.6909 0.49073farmerspositive 0.00317148 0.613108 0.0052 0.99588foodpriceinflation 0.252773 0.190902 1.3241 0.18751foodriots 1.16812 0.344818 3.3876 0.00090 ***hunger 0.523185 0.241844 2.1633 0.03212 **malnutrition -1.22452 0.777857 -1.5742 0.11757poverty 0.403077 0.303807 1.3268 0.18663dummy2008 2.9096 1.1264 2.5831 0.01076 **

Mean dependent var 1.343750 S.D. dependent var 3.585517Sum squared resid 275.8696 S.E. of regression 1.365278R-squared 0.865041 Adjusted R-squared 0.855010F(11, 148) 54.68413 P-value(F) 1.60e-46Log-likelihood -270.6105 Akaike criterion 565.2210Schwarz criterion 602.1231 Hannan-Quinn 580.2057rho 0.093123 Durbin-Watson 1.789403

Model 2 - The Economist: OLS, using observations 2000:01-2013:04 (T = 160)Dependent variable: foodcrisis

HAC standard errors, bandwidth 4 (Bartlett kernel)Coefficient Std. Error t-ratio p-value

const -0.190439 0.0736559 -2.5855 0.01069 **time 0.00417043 0.00134329 3.1046 0.00228 ***africafoodcrisis -0.482105 0.206791 -2.3314 0.02108 **famine 0.182977 0.142861 1.2808 0.20227farmersnegative 0.0635489 0.201982 0.3146 0.75349farmerspositive 0.379364 0.375707 1.0097 0.31427foodpriceinflation 0.0517955 0.0848167 0.6107 0.54235foodriots 0.254857 0.264041 0.9652 0.33601hunger 0.381222 0.15113 2.5225 0.01271 **malnutrition 0.43676 0.795923 0.5487 0.58401poverty 0.294584 0.156199 1.8860 0.06126 *dummy2008 0.618568 0.491453 1.2587 0.21014

26

Mean dependent var 0.556250 S.D. dependent var 1.722025Sum squared resid 70.19006 S.E. of regression 0.688663R-squared 0.851133 Adjusted R-squared 0.840068F(11, 148) 120.9529 P-value(F) 3.92e-68Log-likelihood -161.1128 Akaike criterion 346.2256Schwarz criterion 383.1277 Hannan-Quinn 361.2103rho 0.105323 Durbin-Watson 1.786302

Model 3 - The Mail & Guardian: OLS, using observations 2000:01-2013:04 (T = 160)Dependent variable: foodcrisis

HAC standard errors, bandwidth 4 (Bartlett kernel)Coefficient Std. Error t-ratio p-value

const 0.0229829 0.0402139 0.5715 0.56852time -0.00104558 0.000868051 -1.2045 0.23031africafoodcrisis -0.0564455 0.0699588 -0.8068 0.42105famine -0.216399 0.141467 -1.5297 0.12823farmersnegative -0.100532 0.230262 -0.4366 0.66304farmerspositive 0.762203 0.815289 0.9349 0.35137foodpriceinflation 0.424674 0.117599 3.6112 0.00042 ***foodriots -0.392462 0.160793 -2.4408 0.01584 **hunger 0.605142 0.313041 1.9331 0.05513 *malnutrition -0.721416 0.299938 -2.4052 0.01740 **poverty 0.26651 0.237242 1.1234 0.26310dummy2008 0.656837 0.525929 1.2489 0.21367

Mean dependent var 0.556250 S.D. dependent var 2.610446Sum squared resid 136.8668 S.E. of regression 0.961653R-squared 0.873680 Adjusted R-squared 0.864291F(11, 148) 22.64984 P-value(F) 1.40e-26Log-likelihood -214.5369 Akaike criterion 453.0738Schwarz criterion 489.9759 Hannan-Quinn 468.0585rho -0.112438 Durbin-Watson 2.224685

*A p-value below 0.01 indicates statistical significance at the 1% level and is marked with

***. ** indicates significance between the 1% and 5% levels, and * indicates significance

between the 5% and 10% levels. Model selection statistics (the Akaike Information Criterion or

AIC and Schwarz’s Bayesian Information Criterion) are also given. The formula used for the AIC

is that given by Akaike (1974), namely minus two times the maximized log-likelihood plus two

times the number of parameters estimated.

27