- Email: [email protected]
An empirical analysis of risk in conventional and organic arable farming in The Netherlands
Europ. J. Agronomy 79 (2016) 100–106
Contents lists available at ScienceDirect
European Journal of Agronomy journal homepage: www.elsevier.com/locate/eja
An empirical analysis of risk in conventional and organic arable farming in The Netherlands P.B.M. Berentsen a,∗ , M.A.P.M. van Asseldonk b a b
Business Economics Group, Wageningen University, Hollandseweg 1, 6706 KN, Wageningen, The Netherlands Agricultural Economics Research Institute, Wageningen University and Research Centre, Hollandseweg 1, 6706 KN, Wageningen, The Netherlands
a r t i c l e
i n f o
Article history: Received 10 December 2015 Received in revised form 27 May 2016 Accepted 6 June 2016 Keywords: Arable farming Organic farming Risk assessment Error component implicit detrending method
a b s t r a c t This paper assesses and compares risk in conventional and organic arable farming in The Netherlands with respect to family farm income and underlying price and production variables. To investigate the risk factors the farm accountancy data network was used containing unbalanced panel data from 196 conventional and 29 organic representative Dutch arable farms (for the period 2002 up to and including 2011). Variables with regard to price and production risk were identified using a family farm income analysis scheme. Price risk variables are input and output prices, while yield volatility of different crops is the main production risk variable. To assess risk, an error components implicit detrending method was applied and the resulting detrended standard deviations were compared between conventional and organic farms. Results indicate that the risk at the level of family farm income is higher in organic farming. The underlying variables show higher risk for organic farms in crop yields, crop prices and variable input costs per crop. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Adverse environmental effects of conventional agricultural production systems have increased the demand for more sustainable production systems. Organic farming is recognized as a possible way forward to improve sustainability in agriculture (Tuomisto et al., 2012; Rigby and Caceres, 2001). The focus of organic agriculture on the environment is clearly stated by the European Commission who characterizes organic farming as farming that “relies on a number of objectives and principles as well as common practices, designed to minimise the human impact on the environment, while ensuring the agricultural system operates as naturally as possible” (EU, 2015). From this characterization follow EU production rules about organic farming that have to be respected in order to label the products as organic. For arable farming these are rules as (1) prohibition of the use of synthetic fertilizers and synthetic pesticides and herbicides, (2) the requirement to use only seeding material and propagating material produced organically, and (3) the requirement to apply wide crop rotations. The down-
∗ Corresponding author. E-mail addresses: [email protected] (P.B.M. Berentsen), [email protected] (M.A.P.M. van Asseldonk). http://dx.doi.org/10.1016/j.eja.2016.06.002 1161-0301/© 2016 Elsevier B.V. All rights reserved.
side effects of these rules are that yields in organic production are generally lower than in conventional production while yield variability generally is higher (Stockdale et al., 2001). These downside effects are generally compensated by higher farm gate prices for organic products. This leads to the general impression that organic farming is characterized by higher income levels but also by higher risk levels. Empirical research on comparing risk between conventional and organic arable farming can be divided into studies that take the farm level as the level of analysis (e.g. Tiedemann and Latacz-Lohmann, 2013; Gardebroek et al., 2010; Serra et al., 2008) and studies that focus on crop level (e.g. Palmer et al., 2013; Delmotte et al., 2011). Studies of the former type have in common that outputs are always aggregated on a revenue basis while the level of aggregation of inputs differs between studies. The use of aggregated revenues limits the value of a study as it does not allow for analysing separately price and yield risk, and it also does not allow for risk analysis at crop level. The combination of farm level and crop level risk analysis is highly relevant for arable farming as arable farms are typically set up as diversified multi-commodity operations. While this may partly be based on agronomic grounds, it likewise has important impacts on the farm-level risk exposure. Volatility at farm level is caused by joint volatility in variable input costs, crop yields and output prices of crops cultivated within the farm portfolio. Stud-
P.B.M. Berentsen, M.A.P.M. van Asseldonk / Europ. J. Agronomy 79 (2016) 100–106
ies at crop level focus, for example, on yields and yield variation of potatoes (Palmer et al., 2013) and of rice (Delmotte et al., 2011). As these studies do not take into account price risk, the contribution to an analysis of farm level income risk is limited. The current study tries to combine both approaches in order to be able to draw conclusions about differences in risk between conventional and organic arable farming both at the level of the farm income and at the level of individual crops. The objective of this paper is to compare income risk of conventional and organic arable farms in The Netherlands and to trace back the income risk to production and price risk of the main crops. The basis for this is representative data for The Netherlands from both conventional and organic arable farms over a period of ten years. Risk measured in this paper both in absolute terms, as standard deviation (SD), and in relative terms, as coefficient of variation (CV).
2. Materials and methods 2.1. Description of the data The analysis is based on Dutch arable farm data for the years 2002–2011 recorded via the Farm Accountancy Data Network (FADN) of the Agricultural Economics Research Institute in The Netherlands. FADN consists of an annual survey carried out by the member states of the European Union. To assure representative data, in each member state a random stratified sample is constructed, including around 2% of the farms, and based on three criteria: region, economic size and type of farming. Type of farming indicates the most important agricultural activity or set of activities on a farm, for example arable farming, dairy farming or mixed farming. For arable and dairy farming the type of farming includes both conventional and organic farming systems. The survey is a rotating survey, meaning that the number of years farms are in the survey can differ among farms (i.e. unbalanced panel). The advantage of FADN is that it is a harmonized data source with similar bookkeeping principles in all member states. A further advantage, which is explicitly used in this research, is the micro economic nature of the data source. Detailed information is available of individual farms, which provides the opportunity to conduct analysis at farm level and which gives insight in the distribution and differences in incomes between farms. Furthermore it makes it possible to follow the performance of a farm during consecutive years (Court of Auditors, 2004). The total number of specialised arable farms in the database available for this analysis was 287. Because of the interest in within farm volatility, farms that were only one or two years in the database were excluded from the analysis. Of the remaining farms (225) the majority were conventional farms (196) while 29 were organic. The farms in the database are either conventional or organic for the whole period they are in the sample, so farms in transition between conventional and organic are not included. The average number of years farms were in the in the database is 6.9 for conventional and 5.8 for organic farms. Each individual farm in the sample has a sample weight indicating the number of farms in the total population that are represented by each particular farm. Due to the fact that the strata in the population (used for the stratified sample) are different in size and in homogeneity, weights can differ between sample farms. Due to the continuously changing population (some farms exit, while others expand) even the weight of each individual farm in the sample can change over the years. The database contains both technical and economic variables. A distinction can be made between variables on farm structure, like availability of land (on average 61 ha for conventional and 48 ha
101
for organic farms) and of other assets, and variables indicating how well the farm is managed, ranging from yields per ha for different crops to family farm income. 2.2. Selection of risk variables Risk in arable farming consists mainly of production and price risk of inputs and produced outputs. Production risk of output stems from weather uncertainty affecting crop yield and from inherent uncertainty in crop production (e.g., plant diseases). Price risk of outputs stems from uncertainty in market conditions (e.g. supply and demand) and from uncertainty in product quality which follows for an important part from weather conditions. Also on the input side there can be variation in prices of inputs (and amounts required). At the level of individual crops, however, the database only contains the costs of different inputs, so a distinction between amount and price regarding inputs cannot be made. All these factors together jointly cause income risk. The analysed risk variables in this paper are depicted in Fig. 1. This analysis scheme captures the relation between technical and economic variables and the family farm income. The relations between the variables in the analysis scheme are arithmetic which means that a variable can be calculated from its underlying variables. The grey boxes in Fig. 1 represent the variables which are considered with regard to analysis of price and production risk. Price risk is captured by the variance in crop prices. Production risk is included in the variance in crop yields. Variance in variable cost components captures both price and production risk. With regard to paid labour it should be noted that cost of paid labour is not available per crop but only at the farm level. Fixed cost is excluded from the analysis because this has by definition a low variance, so there is a low risk contribution. 2.3. Method of analysis Differences in risk between conventional and organic farming follow from differences in the within farm standard deviations (SD’s) of the variables described in the previous section. For a proper comparison a farm specific approach is essential, meaning that farm individual data need to be detrended for general observed trends like weather variability (Flaten et al., 2011). A reason for doing so is that the series of individual farm data are not all from the same period. An example may explain this. Suppose the SD’s of yields of a certain crop of two farms are compared in order to determine their risk exposures. Next, suppose the available data of the first farm spans from 2001 to 2005 while those of the second farm spans from 2004 to 2007. Now suppose that 2006 was a very dry year with extremely low crop yield all over the country. This would amplify the SD of production of the second farm while this phenomenon is not specific for that particular farm. This could erroneously lead to the conclusion that in general the SD of the particular crop yield of the second farm is higher than on the first farm. A basic assumption for detrending is that all farms, conventional and organic, are impacted in a similar way by the phenomenon that causes the trends, but the size of the impact might differ between farms. Following Atwood et al. (2003) and Flaten et al. (2011) detrending was done by means of an Error Components Implicit Detrending (ECID) procedure. This includes four successive steps (using wheat yield as an example) to derive time variant farm specific deviations: a) Compute the overall national average wheat yield per ha (Ynat ) in each year (t), Ynat,t ,; b) Compute the yield deviation () of each farm i from the national yield for each year t the farm is in the sample: i,t = Y i,t − Y nat,t (1)
102
P.B.M. Berentsen, M.A.P.M. van Asseldonk / Europ. J. Agronomy 79 (2016) 100–106
Fig. 1. Scheme for analysing family farm income of arable farms. Shadowed boxes show the risk variables included in this research.
where Yi,t is the average wheat yield per ha of farm i in year t;
3. Results
• Compute the average (av) yield deviation over the years for each individual farm:
3.1. Farm level results
i,av = Y i,av − Y nat,av,i (2) Note that the national average is specific for farm i as it only refers to the years farm i is in the sample; • Compute the time variant farm specific deviation (i,t ):
εi,t = Y i,t − Y nat,t − (Y i,av − Y nat,av,i )(3) After computing the time variant farm specific deviation for each of the n years the farm is in the database, the SD per farm can be calculated as:
n ε2i,t
SDi =
3.2. Selection of crops
t=1
(4)
t−1
The average standard deviation for a group of farms (either conventional or organic farms) can be calculated by multiplying each farm by its respective total weight in the sample (wi ) summing up all products and dividing the sum of the products by the sum of the weights: SDaverage =
Table 1 shows the mean values and the standard deviations of the family farm income, the farm gross margin and paid labour costs for the conventional and for the organic farms. The results show that absolute income risk, in terms of SD, is higher on organic farms (P = 0.03). This goes along with a higher mean income on organic farms (P = 0.22). The results for the farm gross margin are very similar to those of the family farm income, meaning a higher farm gross margin for organic farming (P = 0.35) and a higher SD (P = 0.00). For paid labour, the higher mean value on organic farms (P = 0.00) can be explained from the fact that organic farms rely for some crops on hand weeding instead of using chemicals. The difference in SD’s (P = 0.00) shows that this reliance on hand work explains part of the higher income risk of organic farming.
n w ∗ SDi i=1 i n i=1
wi
(5)
Besides means and SD’s, coefficients of variation (CV) are derived to quantify the relative risk exposure in order to test possible differences in risk between conventional and organic farms. The CV of a variable is the variable’s SD divided by its mean value.
The main arable crops in The Netherlands and their appearance on the conventional and organic farms in the sample are shown in Table 2, both in terms of number of farms that grow each crop and in terms of the average area that the crop takes on the farms that grow the crop. From the table it can be concluded that organic arable farming is much more diversified with regard to the number of crops grown on the farms than conventional arable farming. First of all, the percentages of the total number of farms that grow a certain crop are lower on organic farms and second, the area allocated to each crop is lower on organic farms. Of the main arable crops, starch potato is not grown on organic farms, while sugar beet is grown on only one organic farm. The other crops are grown on at least 4 organic farms. The requirement of at least multiple farms growing each crop results in the selection of wheat, seed potato, barley, ware potato, and seed onion as arable crops for further analysis regarding the difference in risk they pose.
P.B.M. Berentsen, M.A.P.M. van Asseldonk / Europ. J. Agronomy 79 (2016) 100–106
103
Table 1 Mean values and average detrended within-farm standard deviations of farm level economic results. Mean value
Family farm income (euro) Farm Gross margin (euro) Paid labour (euro)
Average detrended within farm SD
Conventional
Organic
P-value
Conventional
Organic
P-value
50,160 153,808 8487
60,315 165,060 28,912
0.22 0.35 0.00
40,177 66,679 6598
58,196 89,522 18,219
0.03 0.00 0.00
Table 2 Number and percentage of the total number of farms that grow different arable crops and the mean area devoted to each crops based on the farms that grow the crop. Conventional # of farms Total number of farms Total area Crops: - Sugar beet - Wheat - Seed potato - Barley - Ware potato - Seed onion - Starch potato
Organic % of farms
Area (ha)
196
# of farms
% of farms
61 187 161 114 109 93 88 12
Area (ha)
29
95 82 58 56 47 45 6
13 22 14 10 15 7 19
48 1 19 5 4 22 10 0
3 66 17 14 76 34 0
6 11 4 7 6 8 0
Table 3 Mean values and average detrended within-farm standard deviations of wheat production variables. Mean value
Gross margin (euro/ha) Revenues (euro/ha) Yield (kg/ha) Price (euro/kg) Costs (euro/ha) Planting and seeding Crop protection Fertilizer Other
Average detrended within farm SD
Conventional
Organic
P-value
Conventional
Organic
P-value
923 1389 10,217 0.14 466 89 192 140 45
1144 1464 6675 0.24 320 122 ZO 83 116
0.42 0.12 0.00 0.00 0.00 0.00 – 0.00 0.00
193 191 1696 0.02 74 18 44 39 29
364 346 1890 0.05 95 35 ZO 67 46
0.00 0.00 0.25 0.00 0.00 0.00 – 0.00 0.00
ZO: Zero Observations.
3.3. Crop level results For wheat the gross margin results show that the gross margin level of organically produced wheat is higher (P = 0.42) (Table 3) and there is more gross margin variation in growing wheat organically than conventionally (P = 0.00). The variation of the revenues is 1.8 times as high in organic farming as in conventional farming. This is mainly caused by the price variation which is 2.5 times as high in organic farming as in conventional farming. Yield variation differs only sightly between organic and conventional farming. Also on the cost side, absolute risk is higher in organic farming than in conventional farming (P = 0.00). This higher risk counts for all underlying cost elements, except for crop protection costs. In organic wheat production there are no non-chemical crop protection methods that can be used. From the results it can be said that from an economic point of view organic wheat production is dominated by conventional wheat production as the gross margin is not significantly different while absolute risk is significantly higher in organic wheat production. Moreover, the higher risk is apparent in almost all underlying elements, which makes it difficult to decrease risk by implementing specific risk management tools. A trade-off effect exists between organic seed potato production and conventional production of seed potatoes (Table 4). Again there is significantly more variation in all revenue and cost elements of organic farming, resulting in a gross margin risk which is 2.8 times as high as in conventional farming. But for seed potato there is a
clear compensation for this higher absolute risk which is the 2.2 times higher mean gross margin level. Barley is a worst case crop for organic farming as the gross margin level for organic barley is lower than that of conventional barley (P = 0.08) while variation is 1.7 times as high for organic farming (Table 5). Significantly higher risk factors for organic barley are the farm gate price of barley and the fertilizer costs. Ware potato does much better again for organic arable farming (Table 6). Like with seed potato the higher gross margin risk (P = 0.00) is compensated by a substantially higher gross margin level (P = 0.00). Apart from the yield all underlying revenue and cost elements show a significantly higher absolute risk for organic farming. The picture for seed onions is similar again to that for seed and ware potato (Table 7). A higher gross margin for organic arable farming (P = 0.00) is associated with higher absolute risk (P = 0.00). Remarkable for seed onions is the lower yield variation for organic arable farming (P = 0.05).
3.4. Coefficients of variation The CV’s of the different variables in Fig. 2 provide the overview in terms of relative risk. Relative risk of the farm level income and gross margin is higher in organic farming (although not significantly). Regarding the different crops there are clearly three crops that contribute to the higher relative income risk (i.e. wheat, seed potato, and barley) while ware potato and onion do not contribute
104
P.B.M. Berentsen, M.A.P.M. van Asseldonk / Europ. J. Agronomy 79 (2016) 100–106
Table 4 Mean values and average detrended within-farm standard deviations of seed potato production variables. Mean value
Gross margin (euro/ha) Revenues (euro/ha) Yield (kg/ha) Price (euro/kg) Costs (euro/ha) Planting and seeding Crop protection Fertilizer Other
Average detrended within farm SD
Conventional
Organic
P-value
Conventional
Organic
P-value
4272 6836 32,021 0.21 2563 1298 698 235 329
9489 12551 30,151 0.42 3062 1796 215 138 1062
0.00 0.00 0.28 0.00 0.10 0.01 0.00 0.04 0.00
1414 1431 3795 0.04 426 306 148 89 157
3926 3763 6621 0.08 1352 685 309 125 804
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.05 0.00
Table 5 Mean values and average detrended within-farm standard deviations of barley production variables. Mean value
Gross margin (euro/ha) Revenues (euro/ha) Yield (kg/ha) Price (euro/kg) Costs (euro/ha) Planting and seeding Crop protection Fertilizer Other
Average detrended within farm SD
Conventional
Organic
P-value
Conventional
Organic
P-value
671 1014 6985 0.15 343 80 115 89 59
524 759 5050 0.17 235 113 ZO 24 99
0.08 0.01 0.01 0.09 0.00 0.01 – 0.00 0.03
211 207 1368 0.02 73 16 42 35 27
350 214 1840 0.03 81 23 ZO 54 38
0.02 0.46 0.22 0.00 0.38 0.32 – 0.03 0.09
ZO: Zero Observations. Table 6 Mean values and average detrended within-farm standard deviations of ware potato production variables. Mean value
Gross margin (euro/ha) Revenues (euro/ha) Yield (kg/ha) Price (euro/kg) Costs (euro/ha) Planting and seeding Crop protection Fertilizer Other
Average detrended within farm SD
Conventional
Organic
P-value
Conventional
Organic
P-value
3349 5155 46,060 0.11 1806 783 577 307 136
6288 8039 25,780 0.34 1751 1335 291 102 302
0.00 0.00 0.00 0.00 0.29 0.00 0.00 0.00 0.00
1497 303 6724 0.03 338 170 107 83 113
2850 2851 7141 0.10 500 363 555 128 198
0.00 0.00 0.31 0.00 0.00 0.00 0.00 0.00 0.00
Table 7 Mean values and average detrended within-farm standard deviations of seed onion production variables. Mean value
Gross margin (euro/ha) Revenues (euro/ha) Yield (kg/ha) Price (euro/kg) Costs (euro/ha) Planting and seeding Crop protection Fertilizer Other
Average detrended within farm SD
Conventional
Organic
P-value
Conventional
Organic
P-value
3789 5443 56,405 0.10 1655 640 610 240 164
7630 9959 32,074 0.35 2328 1110 37 144 1036
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
2426 2481 11,538 0.04 310 111 144 87 198
4676 4879 8447 0.13 1115 492 89 104 932
0.00 0.00 0.05 0.00 0.00 0.00 0.00 0.17 0.00
to that. Relative yield risk is significantly higher for all organically produced crops. 4. Discussion Results at the farm level show that both the income level and its variability are higher for organic farming. This is in line with other research (e.g. Acs et al., 2007a; Oude Lansink and Jensma, 2003). Also higher labour input and higher variability of labour input are confirmed in the literature (e.g. Gardebroek et al., 2010;
Serra et al., 2008). This picture of higher level and higher variability is reinforced when looking at gross margin per crop. Regarding physical crop yield the results are somewhat more diverse. The yield level is lower in organic farming for all crops which is widely confirmed in the literature (e.g. Palmer et al., 2013; Delmotte et al., 2011), but yield gaps differ greatly between crops. Yield gaps between conventionally and organically produced crops are crop specific and depend, among others, on the marginal productivity of fertilizer and of crop protection. Also differences in yield variation between conventionally and organically produced crops differ
P.B.M. Berentsen, M.A.P.M. van Asseldonk / Europ. J. Agronomy 79 (2016) 100–106
105
Fig. 2. Average coefficient of variation for family farm income, paid labour and various crop level variables (* means a significant difference at the 0.05 level).
between crops for similar biophysical reasons. The ratio of yield variation and yield, however, shows a very clear picture of higher relative risk for all crops. Of the variable costs, costs of planting and seeding material show both higher mean levels and higher variation. Finally, costs of crop protection and of fertilizer are always lower in organic farming while variation of these costs is generally higher in organic farming.
The farms in the database were either producing conventionally or organically. Farms in transition between were deliberately skipped as these farms would blur the picture of risk in organic farming for two reasons. First, farms in transition have to produce organically for a period of two years before these farms are certified as organic farms. This means these farms face conventional prices for their products in these two years, while yields drop because of changing from conventional to organic farming. Second, farms in
106
P.B.M. Berentsen, M.A.P.M. van Asseldonk / Europ. J. Agronomy 79 (2016) 100–106
transition have to adapt to rather different practices which require a learning period that might lead to extra low yields or higher input costs in the transition phase. As a consequence these farms face a transition period of very low income (see Acs et al., 2007b). The current paper focuses on a comparison of conventional with established organic farms, leaving the transition period aside. The picture of risk with regard to the crops taken into account in this study is less complete for organic farming as it is for conventional farming. This can be seen from the combination of the percentage of farms that grow each crop and the area devoted to the crop on the farms that grow the crop (Table 2). For conventional farming the crops listed in Table 2 cover 83% of the total area while for organic farming this is only 29%. Detrending farm level data was carried out using weighted average data of all farms in the database. As the database is representative for arable farming in The Netherlands, this means detrending is done at the national level. If there would be substantial differences between regions, and if organic farms would at the same time be concentrated in some regions, national detrending could lead to less reliable results (Flaten et al., 2011). However, given the fact that The Netherlands is a small country with little difference between regions it is not very likely that a regional detrending procedure would generate more robust estimates than the current national approach applied. Moreover, conventional and organic crop production are quite similar in the sense that they experience similar effects from exogenous influencing factors like weather. The detrending method can be easily adapted to compare not just within countries but also between different regions within a country. This research presents several descriptive statistics for organic and conventional arable farms, namely averages, time variant within farm specific SD’s and CV’s. Interpretation of differences in averages and in SD has its limitations for measuring risk. A more proper way to measure risk is to compute CV values, which is a relative value. Yet also this indicator only captures part of the risk, and does not quantify for example down-side risks (i.e. lower partial moments), nor does it quantify the cumulative distribution functions (Hardaker et al., 2004). For example, the probability of negative farm income, or probability of certain below target returns can be alternatively used as a risk measure (Berg and Starp, 2006). Although intuitive more appealing, these measures are more data demanding (more observations, especially for organic farming would be required to estimate the tail). Moreover, frequently used risk efficient crop selection models are mostly based on expected value-variance framework and thus linked with current the CV measures. The results of this study show that organic arable farming is more risky than conventional arable farming. From this it could be concluded that farmers that do convert to organic arable farming might be less risk averse than their colleagues who do not convert under the same circumstances. A lower risk aversion of organic farmers is confirmed by other studies.In an empirical study on arable farmers in The Netherlands,Gardebroek (2006) found that on average organic farmers have significantly lower coefficients of
risk aversion than conventional farmers. Also the stochastic risk programming study on arable farming in The Netherlands of Acs et al. (2009) showed that only farmers with a low risk aversion coefficient would convert to organic farming. Finally, Koesling et al. (2004) found that Norwegian organic farmers perceived themselves to be less risk averse than conventional farmers. References Acs, S., Berentsen, P.B.M., De Wolf, M., Huirne, R.B.M., 2007a. Comparison of conventional and organic arable farming systems in The Netherlands by means of bio-economic modelling. Biol. Agric. Hortic. 24, 341–361. Acs, S., Berentsen, P.B.M., Huirne, R.B.M., 2007b. Conversion to organic arable farming in The Netherlands: a dynamic linear programming analysis. Agric. Syst. 94, 405–415. Acs, S., Berentsen, P.B.M., Huirne, R.B.M., 2009. Effect of yield and price risk on conversion from conventional to organic farming. Aust. J. Agric. Resour. Econ. 53, 393–411. Atwood, J., Shaik, S., Watts, M., 2003. Are crop yields normally distributed? A reexamination. Am. J. Agric. Econ. 85, 888–901. Berg, E., Starp, M., 2006. Farm level risk assessment using downside risk measures. In: Contributed Paper for Presentation at the 26th International Conference of the IAAE Gold Coast, Australia, August 12–18, 2006. Court of Auditors, 2004. Special Report No 14/2003 of the Court of Auditors concerning agricultural statistics measurement of farm incomes. Off. J. Eur. Union C45, 1–26. Delmotte, S., Tittonell, P., Mouret, J.C., Hammond, R., Lopez-Ridaura, S., 2011. On farm assessment of rice variability and productivity gaps between organic and conventional cropping systems under Mediterranean climate. Eur. J. Agron. 35, 223–236. EU, 2015. European Commission: Agricultural and Rural Development. (Available at:) http://ec.europa.eu/agriculture/organic/organic-farming/whatorganic en. Flaten, O., Lien, G., Tveteras, R., 2011. A comparative study of risk exposure in agriculture and aquaculture. Acta Agric. Scand. Sect. C: Food Econ. 8, 20–34. Gardebroek, C., Chavez, M.D., Oude Lansink, A., 2010. Analysing production technology and risk in organic and conventional dutch arable farming using panel data. J. Agric. Econ. 61, 60–75. Gardebroek, C., 2006. Comparing risk attitudes of organic and non-organic farmers with a Bayesian random coefficient model. Eur. Rev. Agric. Econ. 33, 485–510. Hardaker, J.B., Huirne, R.B.M., Anderson, J.R., Lien, G., 2004. Coping with Risk in Agriculture. CABI, Wallingford, UK. Koesling, M., Ebbesvik, M., Lien, G., Flaten, O., Valle, P.S., Arntzen, H., 2004. Risk and risk management in organic and conventional cash crop farming in Norway. Acta Agric. Scand. Sect. C: Food Econ. 4, 195–206. Oude Lansink, A.G.J.M., Jensma, K., 2003. Analysing profits and economic behaviour of organic and conventional Dutch arable farms. Agric. Econ. Rev. 4, 19–31. Palmer, M.W., Cooper, J., Tetard-Jones, C., Srednicka-Tober, D., Baranski, M., Eyre, M., Shotton, P.N., Volakakis, N., Cakmak, I., Ozturk, L., Leifert, C., Wilcockson, S.J., Bilsborrow, P.E., 2013. The influence of organic and conventional fertilisation and crop protection practices, preceding crop, harvest year and weather conditions on yield and quality of potato (Solanum tuberosum) in a long-term management trial. Eur. J. Agron. 49, 83–92. Rigby, D., Caceres, D., 2001. Organic farming and the sustainability of agricultural systems. Agric. Syst. 68, 21–40. Serra, T., Zilberman, D., Gil, J.M., 2008. Differential uncertainties and risk attitudes between conventional and organic producers: the case of spanish arable crop farmers. Agric. Econ. 39, 219–229. Stockdale, E.A., Lampkin, N.H., Hovi, M., Keatinge, R., Lennartsson, E.K.M., Macdonald, D.W., Padel, S., Tattersall, F.H., Wolfe, M.S., Watson, C.A., 2001. Agronomic and environmental implications of organic farming systems. Adv. Agron. 70, 261–327. Tiedemann, T., Latacz-Lohmann, U., 2013. Production risk and technical efficiency in organic and conventional agriculture—the case of arable farms in Germany. J. Agric. Econ. 64, 73–96. Tuomisto, H.L., Hodge, I.D., Riordan, P., Macdonald, D.W., 2012. Does organic farming reduce environmental impacts? – a meta-analysis of European research. J. Environ. Manag. 112, 309–320.
Contents lists available at ScienceDirect
European Journal of Agronomy journal homepage: www.elsevier.com/locate/eja
An empirical analysis of risk in conventional and organic arable farming in The Netherlands P.B.M. Berentsen a,∗ , M.A.P.M. van Asseldonk b a b
Business Economics Group, Wageningen University, Hollandseweg 1, 6706 KN, Wageningen, The Netherlands Agricultural Economics Research Institute, Wageningen University and Research Centre, Hollandseweg 1, 6706 KN, Wageningen, The Netherlands
a r t i c l e
i n f o
Article history: Received 10 December 2015 Received in revised form 27 May 2016 Accepted 6 June 2016 Keywords: Arable farming Organic farming Risk assessment Error component implicit detrending method
a b s t r a c t This paper assesses and compares risk in conventional and organic arable farming in The Netherlands with respect to family farm income and underlying price and production variables. To investigate the risk factors the farm accountancy data network was used containing unbalanced panel data from 196 conventional and 29 organic representative Dutch arable farms (for the period 2002 up to and including 2011). Variables with regard to price and production risk were identified using a family farm income analysis scheme. Price risk variables are input and output prices, while yield volatility of different crops is the main production risk variable. To assess risk, an error components implicit detrending method was applied and the resulting detrended standard deviations were compared between conventional and organic farms. Results indicate that the risk at the level of family farm income is higher in organic farming. The underlying variables show higher risk for organic farms in crop yields, crop prices and variable input costs per crop. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Adverse environmental effects of conventional agricultural production systems have increased the demand for more sustainable production systems. Organic farming is recognized as a possible way forward to improve sustainability in agriculture (Tuomisto et al., 2012; Rigby and Caceres, 2001). The focus of organic agriculture on the environment is clearly stated by the European Commission who characterizes organic farming as farming that “relies on a number of objectives and principles as well as common practices, designed to minimise the human impact on the environment, while ensuring the agricultural system operates as naturally as possible” (EU, 2015). From this characterization follow EU production rules about organic farming that have to be respected in order to label the products as organic. For arable farming these are rules as (1) prohibition of the use of synthetic fertilizers and synthetic pesticides and herbicides, (2) the requirement to use only seeding material and propagating material produced organically, and (3) the requirement to apply wide crop rotations. The down-
∗ Corresponding author. E-mail addresses: [email protected] (P.B.M. Berentsen), [email protected] (M.A.P.M. van Asseldonk). http://dx.doi.org/10.1016/j.eja.2016.06.002 1161-0301/© 2016 Elsevier B.V. All rights reserved.
side effects of these rules are that yields in organic production are generally lower than in conventional production while yield variability generally is higher (Stockdale et al., 2001). These downside effects are generally compensated by higher farm gate prices for organic products. This leads to the general impression that organic farming is characterized by higher income levels but also by higher risk levels. Empirical research on comparing risk between conventional and organic arable farming can be divided into studies that take the farm level as the level of analysis (e.g. Tiedemann and Latacz-Lohmann, 2013; Gardebroek et al., 2010; Serra et al., 2008) and studies that focus on crop level (e.g. Palmer et al., 2013; Delmotte et al., 2011). Studies of the former type have in common that outputs are always aggregated on a revenue basis while the level of aggregation of inputs differs between studies. The use of aggregated revenues limits the value of a study as it does not allow for analysing separately price and yield risk, and it also does not allow for risk analysis at crop level. The combination of farm level and crop level risk analysis is highly relevant for arable farming as arable farms are typically set up as diversified multi-commodity operations. While this may partly be based on agronomic grounds, it likewise has important impacts on the farm-level risk exposure. Volatility at farm level is caused by joint volatility in variable input costs, crop yields and output prices of crops cultivated within the farm portfolio. Stud-
P.B.M. Berentsen, M.A.P.M. van Asseldonk / Europ. J. Agronomy 79 (2016) 100–106
ies at crop level focus, for example, on yields and yield variation of potatoes (Palmer et al., 2013) and of rice (Delmotte et al., 2011). As these studies do not take into account price risk, the contribution to an analysis of farm level income risk is limited. The current study tries to combine both approaches in order to be able to draw conclusions about differences in risk between conventional and organic arable farming both at the level of the farm income and at the level of individual crops. The objective of this paper is to compare income risk of conventional and organic arable farms in The Netherlands and to trace back the income risk to production and price risk of the main crops. The basis for this is representative data for The Netherlands from both conventional and organic arable farms over a period of ten years. Risk measured in this paper both in absolute terms, as standard deviation (SD), and in relative terms, as coefficient of variation (CV).
2. Materials and methods 2.1. Description of the data The analysis is based on Dutch arable farm data for the years 2002–2011 recorded via the Farm Accountancy Data Network (FADN) of the Agricultural Economics Research Institute in The Netherlands. FADN consists of an annual survey carried out by the member states of the European Union. To assure representative data, in each member state a random stratified sample is constructed, including around 2% of the farms, and based on three criteria: region, economic size and type of farming. Type of farming indicates the most important agricultural activity or set of activities on a farm, for example arable farming, dairy farming or mixed farming. For arable and dairy farming the type of farming includes both conventional and organic farming systems. The survey is a rotating survey, meaning that the number of years farms are in the survey can differ among farms (i.e. unbalanced panel). The advantage of FADN is that it is a harmonized data source with similar bookkeeping principles in all member states. A further advantage, which is explicitly used in this research, is the micro economic nature of the data source. Detailed information is available of individual farms, which provides the opportunity to conduct analysis at farm level and which gives insight in the distribution and differences in incomes between farms. Furthermore it makes it possible to follow the performance of a farm during consecutive years (Court of Auditors, 2004). The total number of specialised arable farms in the database available for this analysis was 287. Because of the interest in within farm volatility, farms that were only one or two years in the database were excluded from the analysis. Of the remaining farms (225) the majority were conventional farms (196) while 29 were organic. The farms in the database are either conventional or organic for the whole period they are in the sample, so farms in transition between conventional and organic are not included. The average number of years farms were in the in the database is 6.9 for conventional and 5.8 for organic farms. Each individual farm in the sample has a sample weight indicating the number of farms in the total population that are represented by each particular farm. Due to the fact that the strata in the population (used for the stratified sample) are different in size and in homogeneity, weights can differ between sample farms. Due to the continuously changing population (some farms exit, while others expand) even the weight of each individual farm in the sample can change over the years. The database contains both technical and economic variables. A distinction can be made between variables on farm structure, like availability of land (on average 61 ha for conventional and 48 ha
101
for organic farms) and of other assets, and variables indicating how well the farm is managed, ranging from yields per ha for different crops to family farm income. 2.2. Selection of risk variables Risk in arable farming consists mainly of production and price risk of inputs and produced outputs. Production risk of output stems from weather uncertainty affecting crop yield and from inherent uncertainty in crop production (e.g., plant diseases). Price risk of outputs stems from uncertainty in market conditions (e.g. supply and demand) and from uncertainty in product quality which follows for an important part from weather conditions. Also on the input side there can be variation in prices of inputs (and amounts required). At the level of individual crops, however, the database only contains the costs of different inputs, so a distinction between amount and price regarding inputs cannot be made. All these factors together jointly cause income risk. The analysed risk variables in this paper are depicted in Fig. 1. This analysis scheme captures the relation between technical and economic variables and the family farm income. The relations between the variables in the analysis scheme are arithmetic which means that a variable can be calculated from its underlying variables. The grey boxes in Fig. 1 represent the variables which are considered with regard to analysis of price and production risk. Price risk is captured by the variance in crop prices. Production risk is included in the variance in crop yields. Variance in variable cost components captures both price and production risk. With regard to paid labour it should be noted that cost of paid labour is not available per crop but only at the farm level. Fixed cost is excluded from the analysis because this has by definition a low variance, so there is a low risk contribution. 2.3. Method of analysis Differences in risk between conventional and organic farming follow from differences in the within farm standard deviations (SD’s) of the variables described in the previous section. For a proper comparison a farm specific approach is essential, meaning that farm individual data need to be detrended for general observed trends like weather variability (Flaten et al., 2011). A reason for doing so is that the series of individual farm data are not all from the same period. An example may explain this. Suppose the SD’s of yields of a certain crop of two farms are compared in order to determine their risk exposures. Next, suppose the available data of the first farm spans from 2001 to 2005 while those of the second farm spans from 2004 to 2007. Now suppose that 2006 was a very dry year with extremely low crop yield all over the country. This would amplify the SD of production of the second farm while this phenomenon is not specific for that particular farm. This could erroneously lead to the conclusion that in general the SD of the particular crop yield of the second farm is higher than on the first farm. A basic assumption for detrending is that all farms, conventional and organic, are impacted in a similar way by the phenomenon that causes the trends, but the size of the impact might differ between farms. Following Atwood et al. (2003) and Flaten et al. (2011) detrending was done by means of an Error Components Implicit Detrending (ECID) procedure. This includes four successive steps (using wheat yield as an example) to derive time variant farm specific deviations: a) Compute the overall national average wheat yield per ha (Ynat ) in each year (t), Ynat,t ,; b) Compute the yield deviation () of each farm i from the national yield for each year t the farm is in the sample: i,t = Y i,t − Y nat,t (1)
102
P.B.M. Berentsen, M.A.P.M. van Asseldonk / Europ. J. Agronomy 79 (2016) 100–106
Fig. 1. Scheme for analysing family farm income of arable farms. Shadowed boxes show the risk variables included in this research.
where Yi,t is the average wheat yield per ha of farm i in year t;
3. Results
• Compute the average (av) yield deviation over the years for each individual farm:
3.1. Farm level results
i,av = Y i,av − Y nat,av,i (2) Note that the national average is specific for farm i as it only refers to the years farm i is in the sample; • Compute the time variant farm specific deviation (i,t ):
εi,t = Y i,t − Y nat,t − (Y i,av − Y nat,av,i )(3) After computing the time variant farm specific deviation for each of the n years the farm is in the database, the SD per farm can be calculated as:
n ε2i,t
SDi =
3.2. Selection of crops
t=1
(4)
t−1
The average standard deviation for a group of farms (either conventional or organic farms) can be calculated by multiplying each farm by its respective total weight in the sample (wi ) summing up all products and dividing the sum of the products by the sum of the weights: SDaverage =
Table 1 shows the mean values and the standard deviations of the family farm income, the farm gross margin and paid labour costs for the conventional and for the organic farms. The results show that absolute income risk, in terms of SD, is higher on organic farms (P = 0.03). This goes along with a higher mean income on organic farms (P = 0.22). The results for the farm gross margin are very similar to those of the family farm income, meaning a higher farm gross margin for organic farming (P = 0.35) and a higher SD (P = 0.00). For paid labour, the higher mean value on organic farms (P = 0.00) can be explained from the fact that organic farms rely for some crops on hand weeding instead of using chemicals. The difference in SD’s (P = 0.00) shows that this reliance on hand work explains part of the higher income risk of organic farming.
n w ∗ SDi i=1 i n i=1
wi
(5)
Besides means and SD’s, coefficients of variation (CV) are derived to quantify the relative risk exposure in order to test possible differences in risk between conventional and organic farms. The CV of a variable is the variable’s SD divided by its mean value.
The main arable crops in The Netherlands and their appearance on the conventional and organic farms in the sample are shown in Table 2, both in terms of number of farms that grow each crop and in terms of the average area that the crop takes on the farms that grow the crop. From the table it can be concluded that organic arable farming is much more diversified with regard to the number of crops grown on the farms than conventional arable farming. First of all, the percentages of the total number of farms that grow a certain crop are lower on organic farms and second, the area allocated to each crop is lower on organic farms. Of the main arable crops, starch potato is not grown on organic farms, while sugar beet is grown on only one organic farm. The other crops are grown on at least 4 organic farms. The requirement of at least multiple farms growing each crop results in the selection of wheat, seed potato, barley, ware potato, and seed onion as arable crops for further analysis regarding the difference in risk they pose.
P.B.M. Berentsen, M.A.P.M. van Asseldonk / Europ. J. Agronomy 79 (2016) 100–106
103
Table 1 Mean values and average detrended within-farm standard deviations of farm level economic results. Mean value
Family farm income (euro) Farm Gross margin (euro) Paid labour (euro)
Average detrended within farm SD
Conventional
Organic
P-value
Conventional
Organic
P-value
50,160 153,808 8487
60,315 165,060 28,912
0.22 0.35 0.00
40,177 66,679 6598
58,196 89,522 18,219
0.03 0.00 0.00
Table 2 Number and percentage of the total number of farms that grow different arable crops and the mean area devoted to each crops based on the farms that grow the crop. Conventional # of farms Total number of farms Total area Crops: - Sugar beet - Wheat - Seed potato - Barley - Ware potato - Seed onion - Starch potato
Organic % of farms
Area (ha)
196
# of farms
% of farms
61 187 161 114 109 93 88 12
Area (ha)
29
95 82 58 56 47 45 6
13 22 14 10 15 7 19
48 1 19 5 4 22 10 0
3 66 17 14 76 34 0
6 11 4 7 6 8 0
Table 3 Mean values and average detrended within-farm standard deviations of wheat production variables. Mean value
Gross margin (euro/ha) Revenues (euro/ha) Yield (kg/ha) Price (euro/kg) Costs (euro/ha) Planting and seeding Crop protection Fertilizer Other
Average detrended within farm SD
Conventional
Organic
P-value
Conventional
Organic
P-value
923 1389 10,217 0.14 466 89 192 140 45
1144 1464 6675 0.24 320 122 ZO 83 116
0.42 0.12 0.00 0.00 0.00 0.00 – 0.00 0.00
193 191 1696 0.02 74 18 44 39 29
364 346 1890 0.05 95 35 ZO 67 46
0.00 0.00 0.25 0.00 0.00 0.00 – 0.00 0.00
ZO: Zero Observations.
3.3. Crop level results For wheat the gross margin results show that the gross margin level of organically produced wheat is higher (P = 0.42) (Table 3) and there is more gross margin variation in growing wheat organically than conventionally (P = 0.00). The variation of the revenues is 1.8 times as high in organic farming as in conventional farming. This is mainly caused by the price variation which is 2.5 times as high in organic farming as in conventional farming. Yield variation differs only sightly between organic and conventional farming. Also on the cost side, absolute risk is higher in organic farming than in conventional farming (P = 0.00). This higher risk counts for all underlying cost elements, except for crop protection costs. In organic wheat production there are no non-chemical crop protection methods that can be used. From the results it can be said that from an economic point of view organic wheat production is dominated by conventional wheat production as the gross margin is not significantly different while absolute risk is significantly higher in organic wheat production. Moreover, the higher risk is apparent in almost all underlying elements, which makes it difficult to decrease risk by implementing specific risk management tools. A trade-off effect exists between organic seed potato production and conventional production of seed potatoes (Table 4). Again there is significantly more variation in all revenue and cost elements of organic farming, resulting in a gross margin risk which is 2.8 times as high as in conventional farming. But for seed potato there is a
clear compensation for this higher absolute risk which is the 2.2 times higher mean gross margin level. Barley is a worst case crop for organic farming as the gross margin level for organic barley is lower than that of conventional barley (P = 0.08) while variation is 1.7 times as high for organic farming (Table 5). Significantly higher risk factors for organic barley are the farm gate price of barley and the fertilizer costs. Ware potato does much better again for organic arable farming (Table 6). Like with seed potato the higher gross margin risk (P = 0.00) is compensated by a substantially higher gross margin level (P = 0.00). Apart from the yield all underlying revenue and cost elements show a significantly higher absolute risk for organic farming. The picture for seed onions is similar again to that for seed and ware potato (Table 7). A higher gross margin for organic arable farming (P = 0.00) is associated with higher absolute risk (P = 0.00). Remarkable for seed onions is the lower yield variation for organic arable farming (P = 0.05).
3.4. Coefficients of variation The CV’s of the different variables in Fig. 2 provide the overview in terms of relative risk. Relative risk of the farm level income and gross margin is higher in organic farming (although not significantly). Regarding the different crops there are clearly three crops that contribute to the higher relative income risk (i.e. wheat, seed potato, and barley) while ware potato and onion do not contribute
104
P.B.M. Berentsen, M.A.P.M. van Asseldonk / Europ. J. Agronomy 79 (2016) 100–106
Table 4 Mean values and average detrended within-farm standard deviations of seed potato production variables. Mean value
Gross margin (euro/ha) Revenues (euro/ha) Yield (kg/ha) Price (euro/kg) Costs (euro/ha) Planting and seeding Crop protection Fertilizer Other
Average detrended within farm SD
Conventional
Organic
P-value
Conventional
Organic
P-value
4272 6836 32,021 0.21 2563 1298 698 235 329
9489 12551 30,151 0.42 3062 1796 215 138 1062
0.00 0.00 0.28 0.00 0.10 0.01 0.00 0.04 0.00
1414 1431 3795 0.04 426 306 148 89 157
3926 3763 6621 0.08 1352 685 309 125 804
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.05 0.00
Table 5 Mean values and average detrended within-farm standard deviations of barley production variables. Mean value
Gross margin (euro/ha) Revenues (euro/ha) Yield (kg/ha) Price (euro/kg) Costs (euro/ha) Planting and seeding Crop protection Fertilizer Other
Average detrended within farm SD
Conventional
Organic
P-value
Conventional
Organic
P-value
671 1014 6985 0.15 343 80 115 89 59
524 759 5050 0.17 235 113 ZO 24 99
0.08 0.01 0.01 0.09 0.00 0.01 – 0.00 0.03
211 207 1368 0.02 73 16 42 35 27
350 214 1840 0.03 81 23 ZO 54 38
0.02 0.46 0.22 0.00 0.38 0.32 – 0.03 0.09
ZO: Zero Observations. Table 6 Mean values and average detrended within-farm standard deviations of ware potato production variables. Mean value
Gross margin (euro/ha) Revenues (euro/ha) Yield (kg/ha) Price (euro/kg) Costs (euro/ha) Planting and seeding Crop protection Fertilizer Other
Average detrended within farm SD
Conventional
Organic
P-value
Conventional
Organic
P-value
3349 5155 46,060 0.11 1806 783 577 307 136
6288 8039 25,780 0.34 1751 1335 291 102 302
0.00 0.00 0.00 0.00 0.29 0.00 0.00 0.00 0.00
1497 303 6724 0.03 338 170 107 83 113
2850 2851 7141 0.10 500 363 555 128 198
0.00 0.00 0.31 0.00 0.00 0.00 0.00 0.00 0.00
Table 7 Mean values and average detrended within-farm standard deviations of seed onion production variables. Mean value
Gross margin (euro/ha) Revenues (euro/ha) Yield (kg/ha) Price (euro/kg) Costs (euro/ha) Planting and seeding Crop protection Fertilizer Other
Average detrended within farm SD
Conventional
Organic
P-value
Conventional
Organic
P-value
3789 5443 56,405 0.10 1655 640 610 240 164
7630 9959 32,074 0.35 2328 1110 37 144 1036
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
2426 2481 11,538 0.04 310 111 144 87 198
4676 4879 8447 0.13 1115 492 89 104 932
0.00 0.00 0.05 0.00 0.00 0.00 0.00 0.17 0.00
to that. Relative yield risk is significantly higher for all organically produced crops. 4. Discussion Results at the farm level show that both the income level and its variability are higher for organic farming. This is in line with other research (e.g. Acs et al., 2007a; Oude Lansink and Jensma, 2003). Also higher labour input and higher variability of labour input are confirmed in the literature (e.g. Gardebroek et al., 2010;
Serra et al., 2008). This picture of higher level and higher variability is reinforced when looking at gross margin per crop. Regarding physical crop yield the results are somewhat more diverse. The yield level is lower in organic farming for all crops which is widely confirmed in the literature (e.g. Palmer et al., 2013; Delmotte et al., 2011), but yield gaps differ greatly between crops. Yield gaps between conventionally and organically produced crops are crop specific and depend, among others, on the marginal productivity of fertilizer and of crop protection. Also differences in yield variation between conventionally and organically produced crops differ
P.B.M. Berentsen, M.A.P.M. van Asseldonk / Europ. J. Agronomy 79 (2016) 100–106
105
Fig. 2. Average coefficient of variation for family farm income, paid labour and various crop level variables (* means a significant difference at the 0.05 level).
between crops for similar biophysical reasons. The ratio of yield variation and yield, however, shows a very clear picture of higher relative risk for all crops. Of the variable costs, costs of planting and seeding material show both higher mean levels and higher variation. Finally, costs of crop protection and of fertilizer are always lower in organic farming while variation of these costs is generally higher in organic farming.
The farms in the database were either producing conventionally or organically. Farms in transition between were deliberately skipped as these farms would blur the picture of risk in organic farming for two reasons. First, farms in transition have to produce organically for a period of two years before these farms are certified as organic farms. This means these farms face conventional prices for their products in these two years, while yields drop because of changing from conventional to organic farming. Second, farms in
106
P.B.M. Berentsen, M.A.P.M. van Asseldonk / Europ. J. Agronomy 79 (2016) 100–106
transition have to adapt to rather different practices which require a learning period that might lead to extra low yields or higher input costs in the transition phase. As a consequence these farms face a transition period of very low income (see Acs et al., 2007b). The current paper focuses on a comparison of conventional with established organic farms, leaving the transition period aside. The picture of risk with regard to the crops taken into account in this study is less complete for organic farming as it is for conventional farming. This can be seen from the combination of the percentage of farms that grow each crop and the area devoted to the crop on the farms that grow the crop (Table 2). For conventional farming the crops listed in Table 2 cover 83% of the total area while for organic farming this is only 29%. Detrending farm level data was carried out using weighted average data of all farms in the database. As the database is representative for arable farming in The Netherlands, this means detrending is done at the national level. If there would be substantial differences between regions, and if organic farms would at the same time be concentrated in some regions, national detrending could lead to less reliable results (Flaten et al., 2011). However, given the fact that The Netherlands is a small country with little difference between regions it is not very likely that a regional detrending procedure would generate more robust estimates than the current national approach applied. Moreover, conventional and organic crop production are quite similar in the sense that they experience similar effects from exogenous influencing factors like weather. The detrending method can be easily adapted to compare not just within countries but also between different regions within a country. This research presents several descriptive statistics for organic and conventional arable farms, namely averages, time variant within farm specific SD’s and CV’s. Interpretation of differences in averages and in SD has its limitations for measuring risk. A more proper way to measure risk is to compute CV values, which is a relative value. Yet also this indicator only captures part of the risk, and does not quantify for example down-side risks (i.e. lower partial moments), nor does it quantify the cumulative distribution functions (Hardaker et al., 2004). For example, the probability of negative farm income, or probability of certain below target returns can be alternatively used as a risk measure (Berg and Starp, 2006). Although intuitive more appealing, these measures are more data demanding (more observations, especially for organic farming would be required to estimate the tail). Moreover, frequently used risk efficient crop selection models are mostly based on expected value-variance framework and thus linked with current the CV measures. The results of this study show that organic arable farming is more risky than conventional arable farming. From this it could be concluded that farmers that do convert to organic arable farming might be less risk averse than their colleagues who do not convert under the same circumstances. A lower risk aversion of organic farmers is confirmed by other studies.In an empirical study on arable farmers in The Netherlands,Gardebroek (2006) found that on average organic farmers have significantly lower coefficients of
risk aversion than conventional farmers. Also the stochastic risk programming study on arable farming in The Netherlands of Acs et al. (2009) showed that only farmers with a low risk aversion coefficient would convert to organic farming. Finally, Koesling et al. (2004) found that Norwegian organic farmers perceived themselves to be less risk averse than conventional farmers. References Acs, S., Berentsen, P.B.M., De Wolf, M., Huirne, R.B.M., 2007a. Comparison of conventional and organic arable farming systems in The Netherlands by means of bio-economic modelling. Biol. Agric. Hortic. 24, 341–361. Acs, S., Berentsen, P.B.M., Huirne, R.B.M., 2007b. Conversion to organic arable farming in The Netherlands: a dynamic linear programming analysis. Agric. Syst. 94, 405–415. Acs, S., Berentsen, P.B.M., Huirne, R.B.M., 2009. Effect of yield and price risk on conversion from conventional to organic farming. Aust. J. Agric. Resour. Econ. 53, 393–411. Atwood, J., Shaik, S., Watts, M., 2003. Are crop yields normally distributed? A reexamination. Am. J. Agric. Econ. 85, 888–901. Berg, E., Starp, M., 2006. Farm level risk assessment using downside risk measures. In: Contributed Paper for Presentation at the 26th International Conference of the IAAE Gold Coast, Australia, August 12–18, 2006. Court of Auditors, 2004. Special Report No 14/2003 of the Court of Auditors concerning agricultural statistics measurement of farm incomes. Off. J. Eur. Union C45, 1–26. Delmotte, S., Tittonell, P., Mouret, J.C., Hammond, R., Lopez-Ridaura, S., 2011. On farm assessment of rice variability and productivity gaps between organic and conventional cropping systems under Mediterranean climate. Eur. J. Agron. 35, 223–236. EU, 2015. European Commission: Agricultural and Rural Development. (Available at:) http://ec.europa.eu/agriculture/organic/organic-farming/whatorganic en. Flaten, O., Lien, G., Tveteras, R., 2011. A comparative study of risk exposure in agriculture and aquaculture. Acta Agric. Scand. Sect. C: Food Econ. 8, 20–34. Gardebroek, C., Chavez, M.D., Oude Lansink, A., 2010. Analysing production technology and risk in organic and conventional dutch arable farming using panel data. J. Agric. Econ. 61, 60–75. Gardebroek, C., 2006. Comparing risk attitudes of organic and non-organic farmers with a Bayesian random coefficient model. Eur. Rev. Agric. Econ. 33, 485–510. Hardaker, J.B., Huirne, R.B.M., Anderson, J.R., Lien, G., 2004. Coping with Risk in Agriculture. CABI, Wallingford, UK. Koesling, M., Ebbesvik, M., Lien, G., Flaten, O., Valle, P.S., Arntzen, H., 2004. Risk and risk management in organic and conventional cash crop farming in Norway. Acta Agric. Scand. Sect. C: Food Econ. 4, 195–206. Oude Lansink, A.G.J.M., Jensma, K., 2003. Analysing profits and economic behaviour of organic and conventional Dutch arable farms. Agric. Econ. Rev. 4, 19–31. Palmer, M.W., Cooper, J., Tetard-Jones, C., Srednicka-Tober, D., Baranski, M., Eyre, M., Shotton, P.N., Volakakis, N., Cakmak, I., Ozturk, L., Leifert, C., Wilcockson, S.J., Bilsborrow, P.E., 2013. The influence of organic and conventional fertilisation and crop protection practices, preceding crop, harvest year and weather conditions on yield and quality of potato (Solanum tuberosum) in a long-term management trial. Eur. J. Agron. 49, 83–92. Rigby, D., Caceres, D., 2001. Organic farming and the sustainability of agricultural systems. Agric. Syst. 68, 21–40. Serra, T., Zilberman, D., Gil, J.M., 2008. Differential uncertainties and risk attitudes between conventional and organic producers: the case of spanish arable crop farmers. Agric. Econ. 39, 219–229. Stockdale, E.A., Lampkin, N.H., Hovi, M., Keatinge, R., Lennartsson, E.K.M., Macdonald, D.W., Padel, S., Tattersall, F.H., Wolfe, M.S., Watson, C.A., 2001. Agronomic and environmental implications of organic farming systems. Adv. Agron. 70, 261–327. Tiedemann, T., Latacz-Lohmann, U., 2013. Production risk and technical efficiency in organic and conventional agriculture—the case of arable farms in Germany. J. Agric. Econ. 64, 73–96. Tuomisto, H.L., Hodge, I.D., Riordan, P., Macdonald, D.W., 2012. Does organic farming reduce environmental impacts? – a meta-analysis of European research. J. Environ. Manag. 112, 309–320.