**3. Data and methodology**

The sources of risk and their preferred risk management strategies are obtained from faceto-face interviews of 800 farmers, 400 each in the central and northeast regions of Thailand. The central and north-east regions differ in terms of resources, economic development and income distribution. The central region has a farming area of 8.61 million acres or 19.2 per cent of the total farming area. In 2007, the average monthly income per farm in this region is 15,271 baht.(16) The central region is known as the 'rice bowl of Thailand' and more than half of the country's irrigation systems are located in this region known for wet-rice cultivation.(27) In contrast, the north-east region is defined as the 'poorest region' with a long dry season and an annual rainfall that fluctuates widely each year.(13, 18) Approximately 45 per cent of the total farming area in Thailand is located in this region. In 2007, the average monthly income per farm in this region is 8,344 baht.(16)

A smallholder farmer is defined as a farmer who has a farming area less than 30 rai (4.8 ha). Purposive random sampling was employed to classify a particular group of respondents from a certain portion of the population. The sample selection process is as follows. First, the provinces in each region were separated into two main groups: (a) the provinces with large and medium irrigation systems and (b) the provinces in the rain-fed area. Second, purposive sampling was employed to select smallholder farmers in each group. This procedure ensured that the sample covered smallholder farmers of both the irrigated and rain-fed areas in the central and north-east regions.

The information on the sources of risk and risk management strategies perceptions obtained from the respondents using a five-point Likert scale were analyzed in two steps. First, exploratory factor analysis (EFA) was used to capture the information on the interrelationships among the set of variables. This technique enabled the researcher to manage and reduce the number of original variables into a smaller group of new correlation dimensions (factors), which are linear combinations of the original variables.(28, 29) The Kaiser-Meyer-Oklin (KMO) method measured the appropriateness for factor analysis of both data sets. The KMO index varies from 0 to 1, with results of 0.6 or greater suitable for factor analysis. The latent root criterion (eigenvalue > 1) was estimated to identify how many factors in each data set to extract. After the number of factors had been identified, the orthogonal (varimax) rotational method was performed in order to minimize the number of variables that have high loadings on each factor. A factor loading of ± 0.4 was employed as a cut off criterion to determine the inter correlation among the original variables. In addition, Cronbach Alpha was employed to evaluate the internal consistency of each factor.(28)

The relationships between the socioeconomic variables and the perception of risk sources and risk management strategies of the smallholder farmers were also analyzed. Multiple regression was employed to evaluate the influence of farm and farmer characteristics on the smallholder farmers' risk perception and risk management responses. Diagnostic tests were carried out to verify that there was no violation of the multiple regression assumptions. The model specification for the farmer's perception of risk source with socioeconomic variables is postulated as follows:

$$\begin{aligned} S\_i &= b\_0 + b\_1 AGE + b\_2GEN + b\_3 EDUI + b\_4 EXP + b\_5 OFFW + b\_6 FSIZ + b\_7 F\_{10} \\ &+ b\_7 INCM + b\_8 LOC + b\_9 FinC + b\_{10} AHIN + b\_{11} HSIZ + e \end{aligned} \tag{1}$$

The model for risk management responses with socioeconomic variables is given as follows:

$$\begin{aligned} R\_i &= b\_0 + b\_1 AGE + b\_2GEN + b\_3 EDU + b\_4 EXP + b\_5 OFFW + b\_6 FSIZ \\ &+ b\_7 INCM + b\_8 LOC + b\_9 FinC + b\_{10} AHIN + b\_{11} HSZ + e \end{aligned} \tag{2}$$

where:

452 Risk Management – Current Issues and Challenges

**3. Data and methodology** 

adjustment considerations and the regulatory situation.(23)

damage to rice', 'low output prices', 'pests', and 'high input costs'.

monthly income per farm in this region is 8,344 baht.(16)

areas in the central and north-east regions.

Norway perceived 'increasing farm liquidity', 'disease prevention', 'buying farm insurance' and 'cost of production' as the most important strategies used to deal with risk on their farms.(19) On the other hand, New Zealand farmers used a mix of risk management strategies to reduce risk. The strategies varied among the groups of farmers depending on the nature of the product, market structure and conditions, farmer characteristics, dynamic risk

Despite the fact that the evaluation of farmers' risk perceptions and risk management responses are essential to better understand their risk behaviour and managerial decisions, few studies have explicitly investigated awareness of risk among Thai farmers. Akasinha, Ngamsomsuk, Thongngam, Sinchaikul and Ngamsomsuk examined risk perceptions among rice farmers in Payao and Lampang provinces in the northern region.(26) In their study, the Participatory Risk Mapping (PRM) technique was used to elicit sources of risk. The authors' results showed that rice farmers in Payao faced five major sources of risk including 'outbreak of rice disease', 'insects causing damage to rice', 'high input costs', 'flooding', and 'shortage of water supply'. Farmers in Lampang typically faced 'drought', 'insects causing

The sources of risk and their preferred risk management strategies are obtained from faceto-face interviews of 800 farmers, 400 each in the central and northeast regions of Thailand. The central and north-east regions differ in terms of resources, economic development and income distribution. The central region has a farming area of 8.61 million acres or 19.2 per cent of the total farming area. In 2007, the average monthly income per farm in this region is 15,271 baht.(16) The central region is known as the 'rice bowl of Thailand' and more than half of the country's irrigation systems are located in this region known for wet-rice cultivation.(27) In contrast, the north-east region is defined as the 'poorest region' with a long dry season and an annual rainfall that fluctuates widely each year.(13, 18) Approximately 45 per cent of the total farming area in Thailand is located in this region. In 2007, the average

A smallholder farmer is defined as a farmer who has a farming area less than 30 rai (4.8 ha). Purposive random sampling was employed to classify a particular group of respondents from a certain portion of the population. The sample selection process is as follows. First, the provinces in each region were separated into two main groups: (a) the provinces with large and medium irrigation systems and (b) the provinces in the rain-fed area. Second, purposive sampling was employed to select smallholder farmers in each group. This procedure ensured that the sample covered smallholder farmers of both the irrigated and rain-fed

The information on the sources of risk and risk management strategies perceptions obtained from the respondents using a five-point Likert scale were analyzed in two steps. First, exploratory factor analysis (EFA) was used to capture the information on the interrelationships among the set of variables. This technique enabled the researcher to manage and reduce the Si is source of risk i (from factor analysis); Ri is risk management strategy *i* (from factor analysis); *AGE*= 1, if the respondent's age is over 40 years old, 0 otherwise; *GEN*= 1, if the respondent is male, 0 if female;*EDU* = 1, if the highest education of the respondent is high school and higher, 0 if primary school education or less; *EXP*= 1, if the farming experience is over 30 years, 0 otherwise; *OFFW*= 1, if the respondent has off-farm work, 0 if no off-farm work; *FSIZ* is farm size; *INCM* is net farm income; *LOC*= 1, if the respondent's farm is located in central region, 0 if a farm located in north-east region; *FINC*= 1, if farm has a loan, 0 if farm without a loan; *AHIN*= 1, if the annual household income greater than 90,001 baht, 0 otherwise; *HSIZ* is household size; and e is error term.

#### **4. Results and discussion**

#### **4.1. Socioeconomic characteristics of the farmers**

The household and farm characteristics of the central and north-east region farmers are presented in Table 1. Table 1 shows that except for gender, household size and finance used for the farm business, central and north-east region farmers generally differ in terms of personal

and farm characteristics, and income distribution. The age group distribution indicates that the majority of the farmers in both regions were over 40 years old. Around 40 per cent of the northeast region farmers were over 60 years old, whereas 42 per cent of the central region farmers were between 41-50 years old. The age distribution between the farmers in both regions was significantly different with the north-east region farmers more likely to be older than the central region farmers. Nearly half of the farmers in the north-east had been involved in agricultural work for over 40 years which implies that younger farmers are rare especially in the north-east. This may be a result of the rural-to-urban migration problems in Thailand.

Sources of Risk and Risk Management Strategies: The Case of Smallholder Farmers in a Developing Economy 455

*Region Overall* 

*North-east (n=400)* 

*(n=800)* 

*Test of differrence a* 

as purchasing farm equipment, seeds and fertilizers, but the balance was spent on the

*Central (n=400)* 

**Gender** % 0.66 Male 73.3 75.8 74.5 Female 26.8 24.3 25.5 **Age group** % 67.14\*\*\*

41-50 years old 42.0 22.3 32.1 51-60 years old 30.0 30.5 30.3 Over 60 years old 16.3 39.5 27.9 **Marital status** % 12.52\*\*\*

Married 87.5 86.3 86.9

**Highest education** % 17.79\*\*\*

Primary school 69.5 81.8 75.6 Secondary school 23.5 14.0 18.8

**Farming experience** % 105.69\*\*\*

11-20 years 29.3 10.0 19.6 21-30 years 22.5 16.0 19.3 31-40 years 19.5 22.8 21.1 Over 40 years 16.0 44.8 30.4 **Household size** Persons 4.36 4.28 4.32 -0.66 **Total farm size** rai b 21.40 14.80 18.09 -10.10\*\*\* **Land ownership status** % 168.93\*\*\* Owner-self operated 64.8 89.5 77.1 Lease-self operated 29.3 2.0 15.6

farmer's personal and household consumption, for example, food and clothing.

Less than 30 years old 1.5 0.5 1.0 31-40 years old 10.3 7.3 8.8

Single/Never married 4.0 2.0 3.0

De factor relationship 0.8 4.3 2.5 Divorced/separated 7.8 7.5 7.6

Illiterate 3.3 2.0 2.6

Vocational training 2.3 0.8 1.5 Bachelor degree 1.5 1.5 1.5

Less than 10 years 12.8 6.5 9.6

Tenant 0 8.5 4.3 Other 6.0 0 3.0

Less than 10,000 baht 0 1.3 0.6

**Finance farm business** % 0.15 Yes 69.3 68.0 68.6 **Average net farm income**<sup>c</sup> baht 166,450 42,632 104,541 -19.26\*\*\* **Working off-farm** % 43.29\*\*\* Yes 63.3 40.0 51.6 **Annual household income** % 113.16\*\*\*

*Item Unit* 

Around 75 per cent of the farmers in both regions graduated with a primary education and about three per cent were illiterate. The result indicates that the central region farmers had higher levels of education than the north-east farmers (P<0.01). Mustafa argued that the educational level of farmers affected their decision making capacity.(30) A higher educated farmer was expected to perform better than an uneducated farmer in terms of management skills and farm resource allocation to maximize farm profitability.

The average farm size of the farmers in the central region was 21.40 rai (3.42 ha) of which 30 per cent was self-lease operated. In contrast, farmers in the north-east had an average farm size of 14.80 rai (2.37 ha) of which 90 per cent was self owned. This result indicates that the central region farmers hold average farm sizes larger than north-east farmers (P<0.01). This is consistent with the Office of Agricultural Economics who reported that farmers in the central region usually had an average farm size larger than the north-east farmers.(16)

The results for the average net farm income between the farmers in the central and northeast regions were statistically significant at the one per cent level. This result indicated that the average net farm income of the central farmers was larger than for the north-east farmers. In 2008, the central farmers had an average net farm income of 166,445.05 baht/household, whereas the average net farm income of the north-east farmers was only 42,632.80 baht/household.

In addition, approximately 63 per cent of the central region farmers worked off-farm, which was significantly more than for the north-east farmers (P<0.01). The results also showed that central farmers had significantly higher annual household incomes than north-east farmers.

In terms of farmer access to credit, nearly 70 per cent of the farmers in the central and northeast regions had loans and nearly half of them borrowed from the Bank of Agriculture and Agricultural Cooperatives. In addition, eight per cent of the farmers used their own savings to operate their farm business. Only about four per cent had loans from commercial banks. The majority of the farmers obtained short-term loans (see Table 2). This finding supports Limsombunchai, who argued that smallholder farmers in rural Thailand lacked investment funds due to a credit accessibility barrier.(31)

Nearly 50 per cent of the farmers had small debts. Further, 30 per cent of the farmers in the north-east had outstanding debts of less than 30,000 baht during the 2008 crop year. Similarly, 27 per cent of the farmers in the central region had debts between 31,000-50,000 baht. An average of 72.6 per cent of the loans were used in operating the farm business, such as purchasing farm equipment, seeds and fertilizers, but the balance was spent on the farmer's personal and household consumption, for example, food and clothing.

454 Risk Management – Current Issues and Challenges

42,632.80 baht/household.

funds due to a credit accessibility barrier.(31)

and farm characteristics, and income distribution. The age group distribution indicates that the majority of the farmers in both regions were over 40 years old. Around 40 per cent of the northeast region farmers were over 60 years old, whereas 42 per cent of the central region farmers were between 41-50 years old. The age distribution between the farmers in both regions was significantly different with the north-east region farmers more likely to be older than the central region farmers. Nearly half of the farmers in the north-east had been involved in agricultural work for over 40 years which implies that younger farmers are rare especially in the north-east.

Around 75 per cent of the farmers in both regions graduated with a primary education and about three per cent were illiterate. The result indicates that the central region farmers had higher levels of education than the north-east farmers (P<0.01). Mustafa argued that the educational level of farmers affected their decision making capacity.(30) A higher educated farmer was expected to perform better than an uneducated farmer in terms of management

The average farm size of the farmers in the central region was 21.40 rai (3.42 ha) of which 30 per cent was self-lease operated. In contrast, farmers in the north-east had an average farm size of 14.80 rai (2.37 ha) of which 90 per cent was self owned. This result indicates that the central region farmers hold average farm sizes larger than north-east farmers (P<0.01). This is consistent with the Office of Agricultural Economics who reported that farmers in the

The results for the average net farm income between the farmers in the central and northeast regions were statistically significant at the one per cent level. This result indicated that the average net farm income of the central farmers was larger than for the north-east farmers. In 2008, the central farmers had an average net farm income of 166,445.05 baht/household, whereas the average net farm income of the north-east farmers was only

In addition, approximately 63 per cent of the central region farmers worked off-farm, which was significantly more than for the north-east farmers (P<0.01). The results also showed that central farmers had significantly higher annual household incomes than north-east farmers. In terms of farmer access to credit, nearly 70 per cent of the farmers in the central and northeast regions had loans and nearly half of them borrowed from the Bank of Agriculture and Agricultural Cooperatives. In addition, eight per cent of the farmers used their own savings to operate their farm business. Only about four per cent had loans from commercial banks. The majority of the farmers obtained short-term loans (see Table 2). This finding supports Limsombunchai, who argued that smallholder farmers in rural Thailand lacked investment

Nearly 50 per cent of the farmers had small debts. Further, 30 per cent of the farmers in the north-east had outstanding debts of less than 30,000 baht during the 2008 crop year. Similarly, 27 per cent of the farmers in the central region had debts between 31,000-50,000 baht. An average of 72.6 per cent of the loans were used in operating the farm business, such

central region usually had an average farm size larger than the north-east farmers.(16)

This may be a result of the rural-to-urban migration problems in Thailand.

skills and farm resource allocation to maximize farm profitability.



Sources of Risk and Risk Management Strategies: The Case of Smallholder Farmers in a Developing Economy 457

*Central (n=400)* 

4.22 0.910 *(1)* 4.09 0.901 *(1)* 4.34 0.904 *(1)* 3.92\*\*\*

3.82 0.926 *(2)* 3.83 0.861 *(2)* 3.82 0.988 *(2)* -0.11

3.52 1.153 *(3)* 3.70 1.014 *(3)* 3.34 1.252 *(3)* -4.47\*\*\*

3.48 1.080 *(4)* 3.44 0.992 *(4)* 3.53 1.161 *(4)* 1.28

3.47 0.946 *(5)* 3.58 0.965 *(5)* 3.36 0.915 *(5)* -3.35\*\*\*

3.38 1.090 *(6)* 3.38 1.024 *(6)* 3.39 1.154 *(6)* 0.16

3.38 1.345 *(7)* 3.47 1.092 *(7)* 3.29 1.554 *(7)* -1.92\*

3.30 1.097 *(8)* 3.27 1.029 *(8)* 3.32 1.161 *(8)* 0.71

3.11 1.441 *(10)* 3.09 1.372 *(10)* 3.13 1.508 *(10)* 0.44

3.02 1.259 2.95 1.161 3.10 1.347 1.72\*

2.74 1.145 2.56 1.007 2.91 1.245 4.34\*\*\*

2.73 1.106 2.86 1.054 2.60 1.144 -3.28\*\*\*

Excess rainfall 3.27 1.293 *(9)* 3.59 1.017 *(9)* 2.95 1.453 *(9)* -7.16\*\*\*

High level of debt 2.84 1.075 2.90 1.052 2.77 1.095 -1.75\*

*Mean a SD Rank Mean a SD Rank Mean a SD Rank* 

*North-east (n=400)* 

*Test of diff. b*

Table 3 summarizes the results of the most important perceived sources of risk for the farmers in the central and north-east regions. The table shows that marketing risks associated with 'unexpected variability of input prices' and 'unexpected variability of product prices' had the highest and second highest mean scores for sources risk, respectively rated by the farmers in both regions. The SDs of both sources of risk in each group were less than one and this indicates that those sources of risk gained a high level of

consensus among the farmers in both regions.(24)

*(n=800)* 

*Source of risk Overall* 

Unexpected variability of input

Unexpected variability of product prices

Diseases and pests that affect plants and

Thailand's economic and political situation

Changes in national government laws and policies

Natural disasters such as heat, fire, flood, storm

Changes in the world economic and political situation

Deficiency in rainfall causing drought

Problems with hired

Changes in interest

labour

rates

Accidents or problems with health

prices

animals

Changes in

Unexpected variability of yields

a Test of differences of the central and north-east household and farm characteristics based on chi-square and independent *t* test; \* significant at 10%, \*\* significant at 5% and \*\*\* significant at 1%

b 1 rai = 0.16 ha. c Net farm income is based on the 2008 crop year.

Source: Field survey, 2009

**Table 1.** Household and farm characteristics of the farmers in central and north-east Thailand


a Multiple responses

Source: Field survey, 2009

**Table 2.** Financial background of the farmers in central and north-east Thailand

#### **4.2. Farmers' perceptions of sources of risk and risk management strategies**

#### *4.2.1. Sources of risk*

The mean scores of each source of risk were ranked and the standard deviation (SD) was used to indicate the variation in the ratings. Independent sample t-test was employed to compare mean score differences between the farmers in the central and north-east regions.

Table 3 summarizes the results of the most important perceived sources of risk for the farmers in the central and north-east regions. The table shows that marketing risks associated with 'unexpected variability of input prices' and 'unexpected variability of product prices' had the highest and second highest mean scores for sources risk, respectively rated by the farmers in both regions. The SDs of both sources of risk in each group were less than one and this indicates that those sources of risk gained a high level of consensus among the farmers in both regions.(24)

456 Risk Management – Current Issues and Challenges

independent *t* test; \*

a Multiple responses Source: Field survey, 2009

*4.2.1. Sources of risk* 

Source: Field survey, 2009

*Item Unit* 

b 1 rai = 0.16 ha. c Net farm income is based on the 2008 crop year.

**Sources of finance <sup>a</sup>**

**Duration of credit**

**Outstanding loan debt**

**Average percentage of loan used**

*Item* 

Bank of Agriculture and

*Region Overall* 

*North-east (n=400)* 

*Central (n=400)* 

10,001-30,000 baht 0.8 14.3 7.5 30,001-50,000 baht 5.0 16.3 10.6 50,001-70,000 baht 11.0 15.8 13.4 70,001-90,000 baht 11.5 11.0 11.3 More than 90,001 baht 71.8 41.5 56.6 a Test of differences of the central and north-east household and farm characteristics based on chi-square and

significant at 10%, \*\* significant at 5% and \*\*\* significant at 1%

**Table 1.** Household and farm characteristics of the farmers in central and north-east Thailand

*(n=400)* 

Agricultural Cooperative 57.5 34.3 44.6 Cooperatives 23.7 15.9 19.4 Village funds 11.8 25.4 19.4 Personal funds 3.2 12.4 8.3 Commercial bank 6.8 0.3 3.9

Less than 1 year 65.0 72.8 68.9 Greater than 3 years 6.9 20.6 13.7

Under 30,000 baht 14.4 29.4 21.9 31,000-50,000 baht 27.4 23.2 25.3 Over 91,000 baht 13.7 21.0 17.3

On-farm activities 79.8 65.1 72.6 Household expenses 20.1 34.7 27.4

**4.2. Farmers' perceptions of sources of risk and risk management strategies** 

The mean scores of each source of risk were ranked and the standard deviation (SD) was used to indicate the variation in the ratings. Independent sample t-test was employed to compare mean score differences between the farmers in the central and north-east regions.

**Table 2.** Financial background of the farmers in central and north-east Thailand

*(n=800)* 

*Region Overall* 

*Central (n=800)* 

*North-east (n=400)* 

*Test of differrence a* 



Sources of Risk and Risk Management Strategies: The Case of Smallholder Farmers in a Developing Economy 459

'Unexpected variability of yields' was ranked the fifth most important source of risk in both regions. In addition, the financial risks associated with 'changes in interest rates' and 'high

Sources of risk that obtained low mean scores included 'changes in technology and breeding', 'changes in land prices', 'risk from theft', 'changes in the situation of farm

Comparisons of risk perception between the farmers in the central and north-east regions showed significant differences in most sources of risk. This interesting finding might be attributable to the fact that sources of risk vary depending on the farm's geographical condition, farm type, the environmental impact and the country's political and economic situation. Evidently, the small farm business may be affected in different ways by changes in

Table 4 summarizes the results of the perceptions of risk management strategies elicited from the farmers in the central and north-east regions. Production and financial strategies were considered more important managerial responses to risk than marketing strategies by

Among the production strategies perceived by the central region farmers, 'purchase farm machinery to replace labour' was the most important with an average rating of 3.45. Nearly 60 per cent of central region farmers reported using this strategy to cope with hired agricultural labour problems on their farms. From the survey, farm machinery, such as hand tractors and four-wheel tractors, was widely used among the central region farmers. This reflects the imbalance problem between agricultural and industrial labour forces in Thailand. This finding supports Ahmad and Isvilanonda who argued that the rural labour force preferred to work in the industrial sector more than in the agricultural sector due to the gap in wage rates.(33) This may be caused by the lack of agricultural labour especially in

'Storing feed and/or seed reserves' and 'have a farm reservoir for water supplies in dry season' showed significant differences in importance between the farmers in the central and north-east regions (P<0.01). North-east farmers perceived the importance of these two production strategies higher than central region farmers. They rated 'storing feed and/or seed reserves' as the most important production strategies and 'having a farm reservoir for water supplies in dry season' was ranked third with mean ratings of 3.61 and 3.47, respectively. Over 80 per cent of the north-east farmers preferred 'storing feed and/or seed reserves' in managing their small farm operations and approximately 65 per cent of them preferred using the 'having a farm reservoir for water supplies in dry season' strategy on their farm. This indicates that the north-east farmers were confronted with the variability of

'Having diversified crop, animal or other enterprises' and 'planting several varieties of crops' were the least important production strategies for both groups. The north-east

levels of debt' were considered as 'quite important' by all farmers.

families' and 'unable to meet contracting obligations'.

the central region, which has many factories located there.

these sources of risk.

*4.2.2. Risk management strategies* 

the farmers in both regions.

input prices and severe droughts.

a Likert scale is used from 1 (not important) to 5 (extremely important).

b The mean scores of central and north-east farmers are significantly difference at \* *P*<0.1, \*\**P*<0.05 and \*\*\**P*<0.01 based on independent samples *t* test.

Source: Field survey, 2009

**Table 3.** Ranking of perceptions of sources of risk by sampled farmers in central and north-east Thailand

The survey results showed that the uncertainty of input prices and product prices have become increasingly worrying among smallholder farmers in the central and north-east regions. This is probably due to the fact that both sources of risk are out of the farmers' control but directly affect their farm incomes. The prices of the major cash crops in Thailand, such as rice, cassava and sugarcane, are unstable; they depend on supply and demand in both local and international markets. Similarly, the average prices of the major farm inputs such as fertilizer NPK 16-20-0, which is widely used by rice farmers, fluctuated from 9,485 baht/tonne in 2006 to a peak of 19,386 baht/tonne in 2008 and then dropped to 16,199 baht/tonne in 2009.(16)

This finding is consistent with those of Patrick et al., Martin, and Flaten et al. who argued that marketing risks associated with the variability of product and input prices were the most important sources of risk considered by the farmers in their respective study areas.(19, 22, 23)

The production risks related to 'diseases and pests affecting plants and animals', 'excess rainfall' and 'natural disasters such as floods' were ranked third, fourth and sixth, among the farmers in the central region with mean scores of 3.70, 3.59 and 3.47, respectively. The results reflect the heavy floods that inundated the central provinces during September 2008. Following this incident, 100,000 rai (16,000 ha.) of farmland in the central region were damaged.(32)

Institutional risks related to 'changes in Thailand's economic and political situation' and 'changes in national government laws and policies' were ranked third and fourth, among the north-east region farmers, respectively. This finding revealed that smallholder farmers were concerned about the effect of the political conflicts in Thailand on their farm operation.

'Unexpected variability of yields' was ranked the fifth most important source of risk in both regions. In addition, the financial risks associated with 'changes in interest rates' and 'high levels of debt' were considered as 'quite important' by all farmers.

Sources of risk that obtained low mean scores included 'changes in technology and breeding', 'changes in land prices', 'risk from theft', 'changes in the situation of farm families' and 'unable to meet contracting obligations'.

Comparisons of risk perception between the farmers in the central and north-east regions showed significant differences in most sources of risk. This interesting finding might be attributable to the fact that sources of risk vary depending on the farm's geographical condition, farm type, the environmental impact and the country's political and economic situation. Evidently, the small farm business may be affected in different ways by changes in these sources of risk.

#### *4.2.2. Risk management strategies*

458 Risk Management – Current Issues and Challenges

*Source of risk Overall* 

Changes in technology and breeding

prices

Changes in land

Changes in family situation such as marital status, inheritances, etc.

Being unable to meet

independent samples *t* test. Source: Field survey, 2009

baht/tonne in 2009.(16)

contracting obligations

Thailand

*(n=800)* 

a Likert scale is used from 1 (not important) to 5 (extremely important).

b The mean scores of central and north-east farmers are significantly difference at \*

*Central (n=400)* 

2.52 1.089 2.49 0.952 2.55 1.211 0.75

2.47 1.222 2.56 1.241 2.38 1.198 -2.03\*\*

1.98 1.032 2.11 0.966 1.85 1.081 -3.52\*\*\*

1.82 1.046 2.13 1.038 1.52 0.965 -8.50\*\*\*

*Mean a SD Rank Mean a SD Rank Mean a SD Rank* 

Risk from theft 2.19 1.179 2.57 1.144 1.82 1.094 -9.44\*\*\*

**Table 3.** Ranking of perceptions of sources of risk by sampled farmers in central and north-east

The survey results showed that the uncertainty of input prices and product prices have become increasingly worrying among smallholder farmers in the central and north-east regions. This is probably due to the fact that both sources of risk are out of the farmers' control but directly affect their farm incomes. The prices of the major cash crops in Thailand, such as rice, cassava and sugarcane, are unstable; they depend on supply and demand in both local and international markets. Similarly, the average prices of the major farm inputs such as fertilizer NPK 16-20-0, which is widely used by rice farmers, fluctuated from 9,485 baht/tonne in 2006 to a peak of 19,386 baht/tonne in 2008 and then dropped to 16,199

This finding is consistent with those of Patrick et al., Martin, and Flaten et al. who argued that marketing risks associated with the variability of product and input prices were the most important sources of risk considered by the farmers in their respective study areas.(19, 22, 23)

The production risks related to 'diseases and pests affecting plants and animals', 'excess rainfall' and 'natural disasters such as floods' were ranked third, fourth and sixth, among the farmers in the central region with mean scores of 3.70, 3.59 and 3.47, respectively. The results reflect the heavy floods that inundated the central provinces during September 2008. Following this

Institutional risks related to 'changes in Thailand's economic and political situation' and 'changes in national government laws and policies' were ranked third and fourth, among the north-east region farmers, respectively. This finding revealed that smallholder farmers were concerned about the effect of the political conflicts in Thailand on their farm operation.

incident, 100,000 rai (16,000 ha.) of farmland in the central region were damaged.(32)

*North-east (n=400)* 

*P*<0.1, \*\**P*<0.05 and \*\*\**P*<0.01 based on

*Test of diff. b*

> Table 4 summarizes the results of the perceptions of risk management strategies elicited from the farmers in the central and north-east regions. Production and financial strategies were considered more important managerial responses to risk than marketing strategies by the farmers in both regions.

> Among the production strategies perceived by the central region farmers, 'purchase farm machinery to replace labour' was the most important with an average rating of 3.45. Nearly 60 per cent of central region farmers reported using this strategy to cope with hired agricultural labour problems on their farms. From the survey, farm machinery, such as hand tractors and four-wheel tractors, was widely used among the central region farmers. This reflects the imbalance problem between agricultural and industrial labour forces in Thailand. This finding supports Ahmad and Isvilanonda who argued that the rural labour force preferred to work in the industrial sector more than in the agricultural sector due to the gap in wage rates.(33) This may be caused by the lack of agricultural labour especially in the central region, which has many factories located there.

> 'Storing feed and/or seed reserves' and 'have a farm reservoir for water supplies in dry season' showed significant differences in importance between the farmers in the central and north-east regions (P<0.01). North-east farmers perceived the importance of these two production strategies higher than central region farmers. They rated 'storing feed and/or seed reserves' as the most important production strategies and 'having a farm reservoir for water supplies in dry season' was ranked third with mean ratings of 3.61 and 3.47, respectively. Over 80 per cent of the north-east farmers preferred 'storing feed and/or seed reserves' in managing their small farm operations and approximately 65 per cent of them preferred using the 'having a farm reservoir for water supplies in dry season' strategy on their farm. This indicates that the north-east farmers were confronted with the variability of input prices and severe droughts.

> 'Having diversified crop, animal or other enterprises' and 'planting several varieties of crops' were the least important production strategies for both groups. The north-east

farmers considered these two production strategies as 'important' but the central region farmers rated them as 'quite important', which is statistically significant different (P<0.01). The results indicated that the lack of farm resources may affect the diversification performance of the farmers in both groups.

Sources of Risk and Risk Management Strategies: The Case of Smallholder Farmers in a Developing Economy 461

*Central (n=400)* 

2.13 12.4 2.32 12.3 1.95 12.5 -4.59\*\*\*

3.13 48.9 *(9)* 3.08 38.5 *(8)* 3.17 59.3 *(10)* 1.17

2.64 31.3 2.92 42.3 2.36 20.3 -6.30\*\*\*

3.02 42.6 2.98 42.0 *(10)* 3.06 43.3 1.18

*Mean a % b Rank Mean a % b Rank Mean a % b Rank* 

Holding cash 3.41 64.8 *(2)* 3.31 60.0 *(3)* 3.52 69.5 *(2)* 2.98\*\*\* Working off farm 3.28 63.3 *(4)* 3.33 68.8 *(2)* 3.24 57.8 *(8)* -1.07 Reduce debt level 3.27 60.0 *(6)* 3.20 48.5 *(5)* 3.33 71.5 *(7)* 1.73\*

**Table 4.** Ranking of perceptions of risk management strategies by sampled farmers in central and

'Use forward contracts' was the least important marketing strategy considered by most central and north-east regions farmers. Only 10 per cent of the farmers in both regions had used this strategy to manage risk. This suggests that the agricultural production under forward contracts in Thailand is still in its developmental stages and is not popular among the smallholder farmers in rural areas. However, the central region farmers perceived the importance of this marketing strategy significantly more than the north-east farmers

The perceptions of risk responses between the farmers in the central and north-east regions were statistically different in many strategies similar to their perceived sources of risk (see Table 4). The findings from the survey revealed that the smallholder farmers in both regions used a mix of risk strategies to manage and reduce the sources of risk they are confronted with. The findings support Martin, who argued that the farmers' selection criteria for risk management strategies varied depending on farm type, climatic conditions, marketing

In this section, the results of the factor analysis of sources of risk and risk management strategies are discussed. Exploratory factor analysis with varimax orthogonal rotation

*North-east (n=400)* 

*P*<0.1, \*\**P*<0.05 and \*\*\**P*<0.01 based on

*Test of diff. c*

*Source of risk Overall* 

Use forward contracts

Leasing farm machinery

Investing in nonfarm businesses

*Miscellaneous strategies:*  Able to adjust quickly to weather, price and other adverse factors

independent samples *t* test Source: Field survey, 2009

north-east Thailand

**4.3. Factor analysis** 

c

*Financial strategies:*

*(n=800)* 

a Likert scale is used from 1 (not important) to 5 (extremely important). b The percentage of farmers using each risk management strategy.

The mean scores of central and north-east farmers are significantly difference at \*

(P<0.01) with the mean scores of 2.32 and 1.95, respectively.

factors and agricultural rules and regulations.(23)

Financial strategies associated with 'holding cash and easily converted cash assets' and 'working off farm to supplement household income' were considered 'important' by the farmers in the central and north-east regions. Approximately 60 per cent of the farmers in both regions reported that they used these two financial strategies. However, the north-east farmers perceived the importance of 'holding cash and easily converted cash assets' significantly higher than the central region farmers. In addition, 'reduce debt level' was given greater importance by the north-east farmers, whereas 'investing in non-farm businesses' was more important among the central region farmers. In terms of marketing strategies, north-east farmers assigned significantly greater rating scores than central region farmers to 'obtaining market information', 'spread sale over several time period' and 'selection of crop and/or animal varieties with low price variability'.



Sources of Risk and Risk Management Strategies: The Case of Smallholder Farmers in a Developing Economy 461

a Likert scale is used from 1 (not important) to 5 (extremely important).

b The percentage of farmers using each risk management strategy.

c The mean scores of central and north-east farmers are significantly difference at \* *P*<0.1, \*\**P*<0.05 and \*\*\**P*<0.01 based on independent samples *t* test

Source: Field survey, 2009

460 Risk Management – Current Issues and Challenges

performance of the farmers in both groups.

*Source of risk Overall* 

*Production strategies:* 

labour

Purchase farm machinery to replace

Storing feed and/or seed reserves

Apply pests and diseases program

Having diversified crop, animal or other

Have a farm reservoir

enterprises

*Marketing strategies:* 

Planting several varieties of crops

Obtaining market information

Spreading sale over several time periods

Selection of crop and/or animal varieties with low price variability

farmers considered these two production strategies as 'important' but the central region farmers rated them as 'quite important', which is statistically significant different (P<0.01). The results indicated that the lack of farm resources may affect the diversification

Financial strategies associated with 'holding cash and easily converted cash assets' and 'working off farm to supplement household income' were considered 'important' by the farmers in the central and north-east regions. Approximately 60 per cent of the farmers in both regions reported that they used these two financial strategies. However, the north-east farmers perceived the importance of 'holding cash and easily converted cash assets' significantly higher than the central region farmers. In addition, 'reduce debt level' was given greater importance by the north-east farmers, whereas 'investing in non-farm businesses' was more important among the central region farmers. In terms of marketing strategies, north-east farmers assigned significantly greater rating scores than central region farmers to 'obtaining market information', 'spread sale over several time period' and

> *Central (n=400)*

3.44 61.6 *(1)* 3.45 58.8 *(1)* 3.43 64.5 *(5)* -0.26

3.40 60.9 *(3)* 3.20 40.8 *(6)* 3.61 81.0 *(1)* 5.49\*\*\*

3.23 53.9 *(7)* 3.26 53.8 *(4)* 3.19 54.0 *(9)* -0.89

3.06 47.9 *(10)* 2.65 35.5 3.47 60.3 *(3)* 9.40\*\*\*

2.94 33.4 2.84 26.0 3.05 40.8 2.65\*\*\*

2.86 30.0 2.71 19.5 3.01 40.5 3.64\*\*\*

3.27 65.3 *(5)* 3.09 51.8 *(7)* 3.46 78.8 *(4)* 4.89\*\*\*

3.19 41.6 *(8)* 3.01 31.5 *(9)* 3.39 51.8 *(6)* 4.48\*\*\*

2.70 24.8 2.61 21.0 2.79 28.5 2.46\*\*

*Mean a % b Rank Mean a % b Rank Mean a % b Rank* 

*North-east (n=400)* 

*Test of diff. c*

'selection of crop and/or animal varieties with low price variability'.

*(n=800)* 

**Table 4.** Ranking of perceptions of risk management strategies by sampled farmers in central and north-east Thailand

'Use forward contracts' was the least important marketing strategy considered by most central and north-east regions farmers. Only 10 per cent of the farmers in both regions had used this strategy to manage risk. This suggests that the agricultural production under forward contracts in Thailand is still in its developmental stages and is not popular among the smallholder farmers in rural areas. However, the central region farmers perceived the importance of this marketing strategy significantly more than the north-east farmers (P<0.01) with the mean scores of 2.32 and 1.95, respectively.

The perceptions of risk responses between the farmers in the central and north-east regions were statistically different in many strategies similar to their perceived sources of risk (see Table 4). The findings from the survey revealed that the smallholder farmers in both regions used a mix of risk strategies to manage and reduce the sources of risk they are confronted with. The findings support Martin, who argued that the farmers' selection criteria for risk management strategies varied depending on farm type, climatic conditions, marketing factors and agricultural rules and regulations.(23)

#### **4.3. Factor analysis**

In this section, the results of the factor analysis of sources of risk and risk management strategies are discussed. Exploratory factor analysis with varimax orthogonal rotation

was applied to the data using SPSS version 15. Exploratory factor analysis is used to reduce the number of sources of risk and risk management strategies for each group of farmers.

Sources of Risk and Risk Management Strategies: The Case of Smallholder Farmers in a Developing Economy 463

*AS1 AS2 AS3 AS4 AS5 AS6* 

**0.923** 0.091 0.005 0.092 0.134 0.053 0.890

**0.875** 0.064 0.066 0.164 0.030 0.050 0.804

**0.833** 0.220 0.003 0.048 0.179 0.094 0.786

0.009 **0.747** 0.121 0.285 0.042 -0.082 0.663

0.132 **0.616** -0.170 -0.147 -0.127 0.427 0.646

0.141 0.103 0.053 0.071 **0.846** 0.017 0.755

0.131 0.122 0.033 0.046 **0.823** 0.135 0.730

0.077 -0.094 -0.014 0.064 0.115 **0.852** 0.758

0.073 0.104 0.329 0.135 0.071 **0.579** 0.483

27.17 11.46 10.70 7.61 7.55 6.69

27.17 38.63 49.33 56.95 64.49 71.19

*nality* 

*Source of risk Factors a Commu-*

Changes in family situation 0.087 **0.748** 0.097 0.079 0.126 -0.176 0.629

Risk from theft 0.107 **0.700** 0.078 0.203 0.151 0.108 0.583

Changes in land prices 0.315 **0.559** -0.014 0.242 0.107 0.087 0.489 Excess rainfall 0.018 0.050 **0.895** 0.086 0.085 -0.039 0.821 Natural disasters 0.033 0.077 **0.862** -0.056 -0.007 0.190 0.789 Changes in interest rates 0.119 0.261 -0.024 **0.827** 0.065 0.162 0.797 High level of debt 0.169 0.220 0.070 **0.825** 0.064 0.010 0.768

Changes in Thailand's economic and political

Changes in the world economic and political

Changes in national government laws and

Being unable to meet contracting obligations

and contractors

yields

product prices

input prices

explained

Problems with hired labour

Unexpected variability of

Unexpected variability of

Unexpected variability of

Diseases and pests that affect plants and animals

Per cent of total variance

Cumulative per cent of the variance explained

Source: Field survey, 2009

Eigenvalues 4.35 1.83 1.71 1.22 1.21 1.07

Cronbach's Alpha 0.889 0.743 0.776 0.763 0.671 0.426 Number of variables 3 5 2 2 2 2

disaster, AS4=financial situation, AS5=yields and product prices and AS6=input prices.

are deleted from the analysis due to these sources of risk have low communalities.

Factor loadings for an absolute value greater than 0.4 are in **bold**.

a Factors AS1-6 are labelled as AS1=economic and political, AS2=personal and farm business environment, AS3=natural

'Accidents or problems with health', 'deficiency in rainfall causing drought' and 'changes in technology and breeding'

**Table 5.** Varimax rotated factor loadings of sources of risk for all sampled in Thailand farmers (n=800)

situation

situation

policies

### *4.3.1. Sources of risk*

The rotated factor loadings of risk sources for all farmers in the central and north-east regions, obtained from the principal component analysis and a varimax orthogonal rotation, are discussed in this section. The KMO measure of data sufficiency was 0.779 and the Bartlett's Test of Sphericity achieved statistical significance (χ2 = 4927.58, P<0.01), both indicating that the data set was appropriate for factor analysis. However, the preliminary results indicated three sources of risk including 'accidents or problems with health', 'deficiency rainfall' and 'changes in technology or breeding' should be eliminated from the factor analysis because of their low communalities (<0.40).(28) Following this, iteration of varimax orthogonal rotation was performed.

The results are presented in Table 5. Latent root criteria (eigenvalues > 1) were specified for six factors (AS1-6) from the 16 sources of risk variables for all farmers in both regions. These six factors can explain almost 71.2 per cent of the total variance. The Cronbach's Alpha values for factors AS1-5 ranged from 0.671 to 0.899, which exceeded the minimum requirement of 0.6. This demonstrates an adequate reliability among those factors. However, the alpha value was somewhat lower (0.426) for factor AS6. Factors AS1-6 can be labelled in accordance with the significant loading variables that were obtained for each factor and explained as follows:

Factor AS1: this factor is named 'economic and political' because of the relatively high loadings on the sources of risk variables with the changes in Thailand and the world economic and political situations and changes in the government laws and policies that affected the small farm operations.

Factor AS2: this factor incorporates a number of sources of risk related to the farm business environment, including risk from being unable to meet contracting obligations, problems with hired labour, theft and changes in land prices. Moreover, risk from changes in family situation (also as personal risk) loaded highly on this factor. Therefore, this factor is named 'personal and farm business environment'.

Factor AS3: this factor consists of the significant loading of 'excess rainfall' and 'natural disaster'. Factor AS3 is labelled 'natural disaster'.

Factor AS4: this factor can be interpreted as the 'financial situation' because of the high factor loadings on the changes in interest rates and high level of debt.

Factor AS5: this factor is related to the risk from unexpected variability in yields and the unpredictable product prices. Thus, this factor is classified as 'yields and product prices'.

Factor AS6: this factor is labelled 'input prices' because of the highest factor loading of the unexpected variability in input prices in this factor.


Sources of Risk and Risk Management Strategies: The Case of Smallholder Farmers in a Developing Economy 463

a Factors AS1-6 are labelled as AS1=economic and political, AS2=personal and farm business environment, AS3=natural disaster, AS4=financial situation, AS5=yields and product prices and AS6=input prices.

'Accidents or problems with health', 'deficiency in rainfall causing drought' and 'changes in technology and breeding' are deleted from the analysis due to these sources of risk have low communalities.

Factor loadings for an absolute value greater than 0.4 are in **bold**.

Source: Field survey, 2009

462 Risk Management – Current Issues and Challenges

varimax orthogonal rotation was performed.

farmers.

*4.3.1. Sources of risk* 

explained as follows:

affected the small farm operations.

'personal and farm business environment'.

disaster'. Factor AS3 is labelled 'natural disaster'.

unexpected variability in input prices in this factor.

factor loadings on the changes in interest rates and high level of debt.

was applied to the data using SPSS version 15. Exploratory factor analysis is used to reduce the number of sources of risk and risk management strategies for each group of

The rotated factor loadings of risk sources for all farmers in the central and north-east regions, obtained from the principal component analysis and a varimax orthogonal rotation, are discussed in this section. The KMO measure of data sufficiency was 0.779 and the Bartlett's Test of Sphericity achieved statistical significance (χ2 = 4927.58, P<0.01), both indicating that the data set was appropriate for factor analysis. However, the preliminary results indicated three sources of risk including 'accidents or problems with health', 'deficiency rainfall' and 'changes in technology or breeding' should be eliminated from the factor analysis because of their low communalities (<0.40).(28) Following this, iteration of

The results are presented in Table 5. Latent root criteria (eigenvalues > 1) were specified for six factors (AS1-6) from the 16 sources of risk variables for all farmers in both regions. These six factors can explain almost 71.2 per cent of the total variance. The Cronbach's Alpha values for factors AS1-5 ranged from 0.671 to 0.899, which exceeded the minimum requirement of 0.6. This demonstrates an adequate reliability among those factors. However, the alpha value was somewhat lower (0.426) for factor AS6. Factors AS1-6 can be labelled in accordance with the significant loading variables that were obtained for each factor and

Factor AS1: this factor is named 'economic and political' because of the relatively high loadings on the sources of risk variables with the changes in Thailand and the world economic and political situations and changes in the government laws and policies that

Factor AS2: this factor incorporates a number of sources of risk related to the farm business environment, including risk from being unable to meet contracting obligations, problems with hired labour, theft and changes in land prices. Moreover, risk from changes in family situation (also as personal risk) loaded highly on this factor. Therefore, this factor is named

Factor AS3: this factor consists of the significant loading of 'excess rainfall' and 'natural

Factor AS4: this factor can be interpreted as the 'financial situation' because of the high

Factor AS5: this factor is related to the risk from unexpected variability in yields and the unpredictable product prices. Thus, this factor is classified as 'yields and product prices'.

Factor AS6: this factor is labelled 'input prices' because of the highest factor loading of the

**Table 5.** Varimax rotated factor loadings of sources of risk for all sampled in Thailand farmers (n=800)

## *4.3.2. Risk management strategies*

Factor analysis was employed to reduce the risk strategy categories as perceived by the farmers in both the central and north-east regions. The KMO measure of data sufficiency was 0.887. In addition, Bartlett's Test of Sphericity was statistically significant at the one per cent level (χ2 = 3301). This indicates that the data were suitable for factor analysis.

Sources of Risk and Risk Management Strategies: The Case of Smallholder Farmers in a Developing Economy 465

Planting several varieties of crops 0.252 **0.742** 0.093 0.095 0.632

Investing in non-farm investment/business 0.172 -0.001 **0.807** 0.124 0.696

Use forward contracts -0.121 **0.441 0.590** 0.076 0.563 Reduce debt level 0.094 0.117 0.061 **0.787** 0.645

Holding cash and easily converted cash assets 0.117 **0.440** 0.177 **0.559** 0.552

a Factors AR1-4 labelled as AR1=farm production and marketing management, AR2=diversification, AR3=off-farm

**Table 6.** Varimax rotated factor loadings of risk management strategies for all farmers sampled

'Able to adjust quickly to weather, price and other adverse factors' and 'purchase farm machinery to replace of labour'

However, in factor AR3, factor analysis grouped the 'use forward contracts' variable, which is unrelated to the definition of this factor. Therefore, the 'use forward contracts' variable was deleted from factor AR3 and the Cronbach Alpha coefficient slightly improved from 0.642 to 0.697. This result illustrated that factor AR3 had a stronger internal consistency after

**4.4. The association between the farmers' characteristics and source of risk and** 

Multiple regression analysis was employed to investigate the relationship between the farmers' socioeconomic characteristics and the perceptions of sources of risk and risk management strategy components obtained from the factor analysis. The summated scales of sources of risk and risk strategy factors of each group of farmers were summed up and averaged based on the relevant variables in each factor structure and their internal

Eigenvalues 4.69 1.28 1.19 1.01 Per cent of total variance explained 33.48 9.14 8.48 7.24 Cumulative per cent of the variance explained 33.48 42.62 51.09 58.33 Cronbach's Alpha 0.742 0.711 0.642 0.596 Number of variables 5 3 3 3

are deleted from the analysis due to these risk management strategies have low communalities.

Obtaining market information on prices forecast

Having diversified crop, animal or other

Selection of crop and/or animal varieties with

Working off farm to supplement net farm

Leasing farm machinery rather than owning

income and AR4=financial management.

Source: Field survey, 2009

inThailand (n=800)

Factor loadings for an absolute values greater than 0.4 are in **bold**.

'use forward contracts' variable was deleted.

**management perception of risks** 

and trends

enterprises

income

them

low price variability

*Risk management strategy Factors a Communality*

*AR1 AR2 AR3 AR4* 

**0.505** 0.363 0.259 0.280 0.532

0.211 **0.796** -0.030 0.147 0.700

0.387 **0.505** 0.345 -0.039 0.525

0.341 0.058 **0.711** 0.143 0.646

0.164 -0.023 0.111 **0.715** 0.551

The first iteration of factor analysis resulted in the removal of 'able to adjust quickly to weather, price and other adverse factors' and 'purchase farm machinery to replace of labour', because these variables exhibited low communalities. Following this, the second rotation was performed with 14 risk strategies.

The final results of the varimax rotated factor loadings for each risk strategy are documented in Table 6. Factor analysis grouped the 14 risk management strategies into four factors. These four factors explained almost 58.33 per cent of the variance.

With regard to reliability, the Cronbach's Alpha values for factors AR1-3 were 0.742, 0.711 and 0.642, respectively. The alpha value for factor AR4 was 0.596, which is very close to the minimum cut-off level of 0.6. The factors AR1-4 can be named according to each factor structure as follows:

Factor one (AR1): this factor has a relatively high loading of the risk strategy variables related to 'apply pests and diseases programme', 'storing feed and/or seed reserves', 'have a farm reservoir for water supplies in dry season', 'spreading sale over several time period' and 'obtaining market information on prices forecast and trends'. This factor is named 'farm production and marketing management'.

Factor two (AR2): this factor is described as 'diversification' because there were significant loadings of risk strategy variables related to 'having diversified crop, animal or other enterprises', 'planting several varieties of crops' and 'selection of crop and/or animal varieties with low price variability'.

Factor three (AR3): this factor is loaded highly on 'investing in non-farm investment/business' and 'working off farm to supplement net farm income', which represent the influence of off-farm income. Thus, factor three is named 'off-farm income'.

Factor four (AR4): this factor is interpreted as 'financial management', which is concerned with 'reduce debt level', 'leasing farm machinery rather than owning them' and 'holding cash and easily converted cash assets'.


Sources of Risk and Risk Management Strategies: The Case of Smallholder Farmers in a Developing Economy 465


a Factors AR1-4 labelled as AR1=farm production and marketing management, AR2=diversification, AR3=off-farm income and AR4=financial management.

'Able to adjust quickly to weather, price and other adverse factors' and 'purchase farm machinery to replace of labour' are deleted from the analysis due to these risk management strategies have low communalities. Factor loadings for an absolute values greater than 0.4 are in **bold**.

Source: Field survey, 2009

464 Risk Management – Current Issues and Challenges

*4.3.2. Risk management strategies* 

structure as follows:

rotation was performed with 14 risk strategies.

production and marketing management'.

varieties with low price variability'.

cash and easily converted cash assets'.

Have a farm reservoir for water supplies in dry

season

four factors explained almost 58.33 per cent of the variance.

Factor analysis was employed to reduce the risk strategy categories as perceived by the farmers in both the central and north-east regions. The KMO measure of data sufficiency was 0.887. In addition, Bartlett's Test of Sphericity was statistically significant at the one per

The first iteration of factor analysis resulted in the removal of 'able to adjust quickly to weather, price and other adverse factors' and 'purchase farm machinery to replace of labour', because these variables exhibited low communalities. Following this, the second

The final results of the varimax rotated factor loadings for each risk strategy are documented in Table 6. Factor analysis grouped the 14 risk management strategies into four factors. These

With regard to reliability, the Cronbach's Alpha values for factors AR1-3 were 0.742, 0.711 and 0.642, respectively. The alpha value for factor AR4 was 0.596, which is very close to the minimum cut-off level of 0.6. The factors AR1-4 can be named according to each factor

Factor one (AR1): this factor has a relatively high loading of the risk strategy variables related to 'apply pests and diseases programme', 'storing feed and/or seed reserves', 'have a farm reservoir for water supplies in dry season', 'spreading sale over several time period' and 'obtaining market information on prices forecast and trends'. This factor is named 'farm

Factor two (AR2): this factor is described as 'diversification' because there were significant loadings of risk strategy variables related to 'having diversified crop, animal or other enterprises', 'planting several varieties of crops' and 'selection of crop and/or animal

Factor three (AR3): this factor is loaded highly on 'investing in non-farm investment/business' and 'working off farm to supplement net farm income', which represent the influence of off-farm income. Thus, factor three is named 'off-farm income'.

Factor four (AR4): this factor is interpreted as 'financial management', which is concerned with 'reduce debt level', 'leasing farm machinery rather than owning them' and 'holding

Apply pests and diseases program **0.655** -0.035 0.318 0.047 0.533 Storing feed and/or seed reserves **0.651** 0.162 -0.025 0.339 0.565

Spreading sale over several time period **0.618** 0.301 0.183 0.159 0.531

*Risk management strategy Factors a Communality*

*AR1 AR2 AR3 AR4* 

**0.641** 0.288 0.022 0.031 0.495

cent level (χ2 = 3301). This indicates that the data were suitable for factor analysis.

**Table 6.** Varimax rotated factor loadings of risk management strategies for all farmers sampled inThailand (n=800)

However, in factor AR3, factor analysis grouped the 'use forward contracts' variable, which is unrelated to the definition of this factor. Therefore, the 'use forward contracts' variable was deleted from factor AR3 and the Cronbach Alpha coefficient slightly improved from 0.642 to 0.697. This result illustrated that factor AR3 had a stronger internal consistency after 'use forward contracts' variable was deleted.
