PPWNov13- Day 1 - Mockshell and Birner- UHohenheim

Understanding the policy process

and landscape through

discourse analysis

Joe Mockshell & Regina Birner

CGIAR Research Program on Policies, Institutions and Markets (PIM) Workshop:

Approaches and Methods for Policy Process Research

November 18-20, 2013, Washington DC, USA

Agricultural policy context

• Why do African governments adopt their policies?

• Focus on incentive systems

– Bates (1981): Government’s incentive to stay in power

– Van de Walle (2001): Neo-patrimonial state

– Jayne et al. (2002): Incentive dilemma – role of donors

– Olper & Raimondi (2010): Role of regime types

– Swinnen (1994 ) Political support function

• Role of ideologies and values – more recent

– Dutt and Mitra (2010): Role of political ideology

• Research gap & study objective

– Empirical research on role of policy beliefs is missing

– Identify prevailing agricultural policy beliefs

• Case study of Ghana and Uganda

2

Discourse analysis approach

• Discourses through

– Texts (Fairclough 1995; van Dijk 1998), Frames (Shöne & Rein 1994), Story-lines (Hajer 2006)

• Study followed Hajer’s approach of discourse analysis

– Identifies story-lines and metaphors in discourses

– Expressed through language use

• Discourse coalition

– Ensemble of story-lines and actors that utter the story

– Actors that share common story-lines in their discourses

• Policy actors have (van Dijk 1998):

– Positive self-representation of their beliefs

– Negative representation of others beliefs

3

Discourse analysis approach

• In-depth interviews

– Broad initial questions (Roe 1994)

• Challenges affecting the agricultural sector

• Policy instruments to address challenges

• Quantitative approach

– Principal Component Analysis (PCA) & Cluster Analysis

• Novelty of the study

– Application of quantitative analysis to

• identify policy themes and

• discourse coalitions from discourses

4

Research method: sampling

Stakeholder organizations Ghana Uganda

Government agencies (Agriculture policy unit, extension & finance)

7 7

Academic (Agriculture, Agricultural Economics & Political science)

3 4

Think tanks (Research) 1 2

Donor agencies & IFIa 5 6

Non-governmental Organization 4 4

Political party representatives & Parliamentarians 5 4

Interest groups (Civil society Organization) 3 2

Farmers (small & large scale) 2 3

Traditional authorities 1 -

Local government 2 2

Total (67 in-depth interviews) 33 34

aInternational Financial Institutions 5

Data analysis

• In the field

– Memo-writing to identify themes

– ‘Study your emerging data’ (Charmaz 2006)

– ‘Completeness’ & ‘dissimilarity’ (Blee & Taylor 2002)

– Additional actors to be involved and new questions

• After the field work: analysis with Nvivo

– Upload of documents

– Transcription of recorded interviews

– Notes of non-recorded interviews

– Coding of texts

– Identification of metaphors and story-lines

– Inclusion of additional texts (government programs,

donor strategies)

6

Analysis: Nodes & Codes

• Total of 60 transcripts

• Identification of 17

policy themes from

coding

• Binary value to each

policy theme

• 1=Yes if the theme

appeared in the policy

discourses of a

particular respondent

without prompting &

0=No if otherwise)

7

NVivo screenshot

Policy themes (PCA results)

8

Ghana Uganda

Policy themes Eigenvalues Policy themes Eigenvalues

Input subsidy & appropriate

technology 5.49

Market price support &

storage 5.82

Quality of agricultural input 2.42

Appropriate technology 2.16

Low agricultural

productivity

1.80

Quality of agricultural

input 1.76

Market price support &

storage

1.46

Inadequate land

policies 1.34

Youth involvement in

agriculture 1.31

Input subsidy & low

agric. Productivity 1.13

Rainfall dependent

farming & crop insurance

1.04

Discourse coalition (Cluster analysis)

Policy stakeholder organizations

Ghana Uganda

Domestic

coalition

Donor

coalition

Domestic

coalition

Donor

coalition

Government agencies

(agricultural policy unit,

extension & finance)

7 0 6 0

Academic (agricultural

economics, political science) 0 3 2 2

Think tank (research) 0 1 2 0

International finance Institutions/

Donor organizations 0 5 0 6

Non-governmental organizations 3 2 3 1

Political party representatives

and parliamentarians 5 0 4 0

Interest groups 3 0 2 0

Local governments 1 0 2 0

Total 19 11 21 9

9

Summary of some key discourses

10

Domestic policy coalition

Public sector centered

Donor-led coalition

Private sector centered

Framing of the main problem

Farmers do not have access to inputs

Institutions are not available, poor implementation, capacity

Views on input subsidies Important to transform agriculture

Subsidies not sustainable, kill private sector initiatives

Views on appropriate technology (mechanization)

Important to modernize agriculture to attract the youth

Should be achieved by the private sector

Self-image

Capable of understanding domestic problems and determining the best policy option for the local economy

Capable of bringing external experience and superior knowledge to provide evidence based policy options

Other-image They come with policies that create dependency

They lack capacity, they do not have any figures to show

Policy beliefs about mechanization

The youth prefers jobs outside

agriculture, which offer “better

jobs than the drudgery that the

youth go to face when they go

into farming, because farming

in Ghana is still largely

dependent on hoe & cutlasses,

so it is a lot of drudgery

involved, so it is not attractive.”

(Member of Ghana Parliament)

“the tractors have a political

image, because they are

big, when they say we have

brought in tractors, when

they say we have brought

in 1000 tractors, you can

make a big political

statement of it.”

(Development consultant)

Domestic coalition beliefs Donor coalition beliefs

• Two worlds in agricultural policy making

• Pay attention to role of policy beliefs in agricultural policy making

11

Thank you

Transform

qualitative

data

Policy

beliefs &

discourse

coalition

PCA &

cluster

analysis

In-depth

Interviews

Policy

themes

coding

of

transcripts

12

Appendix 1: Ghana

PCA: Policy themes, (Pattern Matrix)

Policy themes

Input

subsidy &

appropriate

technology

Quality

of agric.

input

Low agric.

productivity

Market

price

support &

storage

Youth

involveme

nt in

agriculture

Rainfall

dependent

farming &

crop

insurance

Appropriate technology 0.88 0.00 -0.23 -0.17 -0.14 0.15

Inadequate access to credit 0.78 0.01 0.30 0.17 0.04 0.13

Fertiliser subsidy 0.69 -0.14 -0.06 0.24 -0.11 -0.19

Agric. mechanisation 0.51 0.00 0.01 0.21 -0.49 -0.14

Rainfall dependent agric. 0.39 -0.46 -0.20 0.16 -0.07 -0.31

Public extension 0.35 0.04 -0.07 0.72 0.16 0.14

Counterfeit inputs -0.07 0.86 -0.04 -0.13 -0.16 0.01

Quality standards 0.15 0.76 0.03 0.30 0.23 -0.20

Hoe and cutlass farming 0.06 0.39 0.04 0.27 -0.68 -0.17

Irrigation facilities 0.27 -0.37 0.33 0.38 -0.06 -0.17

Low agric. Productivity -0.07 -0.01 -0.89 0.26 0.16 0.10

Private extension -0.17 0.00 0.87 0.06 0.10 0.21

Lack of storage facility -0.07 0.00 -0.02 0.86 -0.20 -0.18

Market price support -0.06 0.01 -0.13 0.86 -0.10 0.10

Youth in agric. policy 0.19 -0.12 0.06 -0.06 -0.79 -0.02

Inadequate land policies -0.19 -0.21 -0.01 0.27 -0.58 0.58

Crop insurance 0.23 -0.04 0.09 -0.01 0.11 0.89

Eigenvalues 5.49 2.42 1.80 1.46 1.31 1.04

13

Appendix 2: Uganda

PCA: Policy themes, (Pattern Matrix)

Policy themes Market

price

support &

storage

Appropriate

technology

Quality of

agric.

input

Inadequate

land

policies

Input subsidy

& low

agricultural

productivity

Quality standards 0.86 -0.20 -0.02 0.08 -0.04

Lack of storage facility 0.80 0.07 -0.08 0.02 0.03

Hoe and cutlass farming 0.59 0.54 -0.13 -0.28 0.14

Market price support 0.50 0.36 -0.09 0.03 -0.19

Youth in agriculture policy 0.45 -0.08 0.25 -0.33 -0.36

Public extension 0.39 0.10 -0.06 0.29 -0.60

Rainfall dependent agric. 0.07 0.90 0.02 0.07 -0.04

Irrigation facilities -0.05 0.88 0.06 -0.06 0.13

Fertiliser subsidy -0.15 0.57 0.19 0.10 -0.45

Counterfeit inputs 0.06 0.42 -0.54 0.40 0.02

Agricultural mechanisation 0.21 0.40 -0.04 -0.31 -0.46

Private extension -0.16 0.30 0.84 0.09 -0.10

Crop insurance 0.08 -0.05 0.65 0.19 0.53

Inadequate land policies -0.05 0.05 -0.10 -0.87 0.01

Appropriate technology 0.26 -0.15 0.22 0.08 -0.77

Low agricultural productivity -0.11 -0.05 -0.06 -0.10 -0.71

Inadequate access to credit 0.24 0.28 -0.18 -0.02 -0.64

Eigenvalues 5.82 2.16 1.76 1.34 1.13

14