The Agricultural Sector

Abstract

This chapter contrasts Timor’s record with the agricultural dynamism evident in much of Southeast Asia. The agricultural sector is low in production, and it is not competitive internationally. Imports have to be resorted to in order to ensure food security. Production rests on shifting cultivation, bush fallow rotation, and simple hand tools. The main crops—rice and maize—both display considerable fluctuations, and their absolute yields are among the lowest in the region. The production of coffee, the most important export crop, is decreasing. The quality of the land is low and possibly declining, capital inputs are sparsely used, and credit is difficult to obtain. Most of the time, unimproved seed is used, but introduction of high-yielding varieties has proved successful for several crops. The low educational level and the lack of an effective extension service make the introduction of innovations difficult.

Appendix: A Model of the Agricultural Household in East Timor

The simple model of farm household behavior below intends to shed some light on the question of whether agricultural ‘entrepreneurs’ will seize the opportunities offered by the market.²¹² In this model, households produce food and a cash crop (which they do not consume themselves) on a given land area. The consumption and production decisions are hence interdependent. The households maximize their utility given the relative prices of cash crops, food crops and purchased goods.²¹³ (The price of purchased goods is used as the numéraire, that is, the prices of other goods can be measured in terms of the number of units of purchased goods they can buy.) Utility maximization yields an equilibrium in which the households distribute the available time between production and leisure, the time devoted to production between cash crops and food production, and the resulting income between the consumption of food and purchased goods, all according to their preferences and the prevailing relative prices. It also determines how much food is sold on the market.

The utility function of the agricultural household is

$$ \mathrm{U}=\mathrm{U}\left(\mathrm{F},\mathrm{M},{\mathrm{L}}_{\mathrm{e}}\right) $$

(5.1)

where $\mathrm{F}$ denotes food, $\mathrm{M}$ purchased goods and ${\mathrm{L}}_{\mathrm{e}}$ leisure.²¹⁴ Utility is positive and diminishing for all arguments of the function.

Food and cash crops are produced using labor (and a fixed quantity of land):

$$ {\mathrm{q}}_{\mathrm{f}}={\mathrm{q}}_{\mathrm{f}}\left({\mathrm{L}}_{\mathrm{f}}\right) $$

(5.2)

$$ {\mathrm{q}}_{\mathrm{c}}={\mathrm{q}}_{\mathrm{c}}\left({\mathrm{L}}_{\mathrm{c}}\right) $$

(5.3)

These production functions display diminishing returns to labor. The total quantity of labor (time) is divided between leisure, food, and cash crop production:

$$ \mathrm{L}={\mathrm{L}}_{\mathrm{e}}+{\mathrm{L}}_{\mathrm{f}}+{\mathrm{L}}_{\mathrm{c}} $$

(5.4)

Households maximize their utility subject to (5.4) and to

$$ {\mathrm{p}}_{\mathrm{m}}\mathrm{M}={\mathrm{p}}_{\mathrm{f}}\left[{\mathrm{q}}_{\mathrm{f}}\left({\mathrm{L}}_{\mathrm{f}}\right)\hbox{--} \mathrm{F}\right]+{\mathrm{p}}_{\mathrm{c}}{\mathrm{q}}_{\mathrm{c}}\left({\mathrm{L}}_{\mathrm{c}}\right) $$

(5.5)

where the p denotes prices. What is available for spending on purchased goods is the sum of the value of food production, minus what is consumed by the household itself, and the value of cash crop production.

Utility maximization yields the following first-order conditions:

$$ {\mathrm{U}}^{\mathrm{M}}\left({\mathrm{p}}_{\mathrm{f}}/{\mathrm{p}}_{\mathrm{m}}\right)={\mathrm{U}}^{\mathrm{F}} $$

(5.6)

$$ {\mathrm{U}}^{\mathrm{L}\mathrm{e}}={\mathrm{U}}^{\mathrm{M}}\left({\mathrm{p}}_{\mathrm{f}}/{\mathrm{p}}_{\mathrm{m}}\right){{\mathrm{q}}_{\mathrm{f}}}^{\mathrm{L}}={\mathrm{U}}^{\mathrm{M}}\left({\mathrm{p}}_{\mathrm{c}}/{\mathrm{p}}_{\mathrm{m}}\right){{\mathrm{q}}_{\mathrm{c}}}^{\mathrm{L}} $$

(5.7)

where the superscripts denote first derivatives. In equilibrium, the valuation of a unit of food must be the same at the margin regardless of whether the household chooses to sell it to buy purchased goods or to consume it in the household. Also, at the margin the household must obtain the same satisfaction of a unit of time regardless of whether it is dedicated to leisure, food production, or cash crop production.

The situation is illustrated in Figure 5.6. Beginning with the $\mathrm{OA}$ curve, this shows the income-leisure tradeoff for the household. The number of hours available is $\mathrm{OB}.$ Work (as opposed to leisure) is measured leftward in the diagram and income downward. The more hours the household works, the higher is its income, which increases at a decreasing rate. Given the preferences of the household for income and leisure, respectively, it chooses point $\mathrm{C}$ on the tradeoff, working $\mathrm{OD}$ hours and taking out $\mathrm{DB}$ hours of free time. This yields a total income of $\mathrm{OK},$ measured in terms of purchased goods.

Fig. 5.6

The equilibrium of the farm household

The $\mathrm{OD}$ hours worked must be distributed between food and cash crop production. If only the latter is chosen, the maximum output will be $\mathrm{OE},$ one of the two endpoints on the household’s production possibility curve. If the household specializes in food production instead, the maximum output is given by the other endpoint of the curve, $\mathrm{OF}.$ The production possibility curve is drawn on the assumption of a given factor endowment and shows the different possible combinations of food and cash crops that can be obtained with the aid of this endowment. The actual choice of production is determined by the relative price of the two crops, as given by the price line $\mathrm{IJ}.$ This line is tangent to the production possibility curve at $\mathrm{G},$ which is where the value of output is maximized at the given prices.

The output combination given by $\mathrm{G}$ gives a farm household income of $\mathrm{OI},$ measured in terms of cash crops, or $\mathrm{OJ},$ if we choose to measure in terms of food instead. Given the relative price prevailing between food and purchased goods, this translates into an income of $\mathrm{OK},$ in terms of the latter, our numéraire. The only thing that remains to be determined is the consumption of the household, given this income, represented by the budget line $\mathrm{JK}.$ Given the preferences of the household for the two goods, the consumption point is $\mathrm{L},$ where $\mathrm{ML}$ units of the purchased goods are consumed and $\mathrm{OM}$ is units of food. This, in turn, leaves a food surplus of $\mathrm{MH},$ which is sold in the market.

Farm Household Response to Market Incentives

Next, we use our model to examine the proposition offered by John Conroy: that Timorese farmers may be expected to increase production for any of the following three reasons²¹⁵:

• 1 increased prices of cash crops,

• 2 increased prices of food crops, and

• 3 reduced prices (increased availability) of goods not manufactured by the farmers but purchased from outside.

In the present chapter, we do not make any assumptions as to where the price changes come from, but they are likely to be intimately connected with, for example, infrastructural development. For example, better roads are likely to increase the net price fetched by farmers both for cash and for food crops while at the same time they should have to pay less for what they buy from the outside.

We next examine the probable effects of each one of these price changes with the aid of our model.

Increased Prices of Cash and Food Crops

Let us begin with what happens when the price of cash crops increases $\left({\mathrm{dp}}_{\mathrm{c}}>0\right)$ while the other two prices are kept constant. The price increase serves to rotate the income-leisure tradeoff in Figure 5.6 anti-clockwise, since the value of the marginal product of labor in the production of cash crops increases. For any given labor effort in agriculture, household income will increase. At the same time, leisure will become more expensive in terms of income forgone. We have an income effect and a substitution effect, and it will not be possible to reach a general conclusion of whether more hours will be devoted to agriculture than before or the number of leisure hours will increase. However, if leisure is an inferior good, as hinted by Conroy, the number of hours worked will increase. This is a quite realistic case in a poor economy. Households have plenty of idle time. Should income possibilities improve, it is quite likely that the farm household prefers to work more and not less. Leisure does not even have to be inferior. It is enough that the substitution effect is strong enough to swamp the increased demand for leisure at given prices.

What will happen to the production of the two goods? The first-order condition (5.7) may be rewritten as:

$$ \left({\mathrm{p}}_{\mathrm{f}}/{\mathrm{p}}_{\mathrm{m}}\right){{\mathrm{q}}_{\mathrm{f}}}^{\mathrm{L}}=\left({\mathrm{p}}_{\mathrm{c}}/{\mathrm{p}}_{\mathrm{m}}\right){{\mathrm{q}}_{\mathrm{c}}}^{\mathrm{L}} $$

(5.8)

Differentiating (5.4) yields:

$$ {\mathrm{dL}}_{\mathrm{f}}=-\left({\mathrm{dL}}_{\mathrm{c}}+{\mathrm{dL}}_{\mathrm{e}}\right) $$

(5.9)

and, from (5.8), allowing ${\mathrm{p}}_{\mathrm{c}}$ to change and keeping ${\mathrm{p}}_{\mathrm{m}}$ and ${\mathrm{p}}_{\mathrm{f}}$ constant (normalizing all three prices = 1, at the outset), we obtain

$$ {{\mathrm{q}}_{\mathrm{f}}}^{\mathrm{L}\mathrm{L}}{\mathrm{dL}}_{\mathrm{f}}={{\mathrm{q}}_{\mathrm{c}}}^{\mathrm{L}\mathrm{L}}{\mathrm{dL}}_{\mathrm{c}}+{{\mathrm{q}}_{\mathrm{c}}}^{\mathrm{L}}{\mathrm{dp}}_{\mathrm{c}} $$

(5.10)

Combining (5.9) and (5.10) gives us

$$ -\left({{\mathrm{q}}_{\mathrm{f}}}^{\mathrm{L}\mathrm{L}}+{{\mathrm{q}}_{\mathrm{c}}}^{\mathrm{L}\mathrm{L}}\right){\mathrm{dL}}_{\mathrm{c}}={{\mathrm{q}}_{\mathrm{c}}}^{\mathrm{L}}{\mathrm{dp}}_{\mathrm{c}}+{{\mathrm{q}}_{\mathrm{f}}}^{\mathrm{L}\mathrm{L}}{\mathrm{dL}}_{\mathrm{e}}>0 $$

(5.11)

provided that ${\mathrm{dL}}_{\mathrm{e}}<0,$ since ${\mathrm{dp}}_{\mathrm{c}}>0,$ and the second derivatives of the production functions are negative. By the same token ${\mathrm{dL}}_{\mathrm{c}}$ must be positive. If the rise in the price of cash crops leads to less leisure, the production of these crops must increase. Even if ${\mathrm{dL}}_{\mathrm{e}}$ is positive, the production of cash crops may increase, if the price rise is large. The production of food crops may increase as well, in the case where the reduction of the hours devoted to leisure exceeds the increase in the number of hours devoted to cash crop production. Finally, what happens to the marketed surplus of food crops is uncertain. The on-farm demand for food will increase if the production of cash crops, and hence income, increases, since the relative price of food and purchased goods remains constant, but food production may also increase.

In the same way as above, it may be demonstrated that an increase in the price of food will lead to increased food production, provided that leisure is reduced or provided that the increase in the price of food is large. What will happen to on-farm consumption and the marketed surplus of food is, however, not clear. When incomes rise in terms of the purchased goods, the farm households will demand more of both these goods and food at given relative prices. However, there is also a substitution effect due to the increase of the relative price of food. This serves to reduce the consumption of food and increase that of purchased goods. The net outcome of this is unclear and will depend on the income elasticity of demand for food products among the farm households themselves.

Normally, we should expect Engel’s Law to operate, that is, when incomes rise, less and less of the income increase will be devoted to food consumption, so that the share of the household budget spent on food will decrease as income increases. In Timor-Leste, however, this is not necessarily the case. Various empirical investigations have shown that households experience a shortage of food between November and April, that is, for some four months every year (cf. Chapter 3). Hence, given the low-income level and the large share of the population living in poverty, it is far from certain that Engel’s Law is operative. It is not possible to rule out the possibility that an increase in the price of food will lead to a reduction of the marketed surplus.

Cheaper Purchased Goods

Finally, we investigate whether a cheapening of the goods that the farm households purchase from the outside will serve as a stimulus to increased production and sales (in the case of food). Again, we differentiate (5.8) but this time holding ${\mathrm{p}}_{\mathrm{f}}$ and ${\mathrm{p}}_{\mathrm{c}}$ constant, while allowing ${\mathrm{p}}_{\mathrm{m}}$ to fall. This yields

$$ {{\mathrm{q}}_{\mathrm{f}}}^{\mathrm{L}\mathrm{L}}{\mathrm{dL}}_{\mathrm{f}}-{{\mathrm{q}}_{\mathrm{f}}}^{\mathrm{L}}{\mathrm{dp}}_{\mathrm{m}}={{\mathrm{q}}_{\mathrm{c}}}^{\mathrm{L}\mathrm{L}}{\mathrm{dL}}_{\mathrm{c}}-{{\mathrm{q}}_{\mathrm{c}}}^{\mathrm{L}}{\mathrm{dp}}_{\mathrm{m}} $$

(5.12)

In equilibrium, the value of the marginal product of labor in food production must equal that in cash crop production (remembering that we have normalized all prices = 1 at the outset). Hence the second terms on both sides of the equality sign in (5.12) are equal. What remains is

$$ {{\mathrm{q}}_{\mathrm{f}}}^{\mathrm{LL}}{\mathrm{dL}}_{\mathrm{f}}={{\mathrm{q}}_{\mathrm{c}}}^{\mathrm{LL}}{\mathrm{dL}}_{\mathrm{c}} $$

(5.13)

Either the production of food and cash crops will both increase or both decrease. Rewriting (5.9) as

$$ {\mathrm{dL}}_{\mathrm{e}}=-\left({\mathrm{dL}}_{\mathrm{f}}+{\mathrm{dL}}_{\mathrm{c}}\right) $$

(5.14)

shows that the crucial determinant here is what happens to leisure as purchased goods become cheaper. Since we have chosen the price of purchased goods as our numéraire, making these goods cheaper amounts to increasing farm household income. The income-leisure tradeoff in Figure 5.6 will rotate anti-clockwise, exactly when the price of cash crops or food increases. Once more, we have an income effect and a substitution effect, and if the net outcome of these is to reduce leisure, the production possibility curve will shift outward and the production of both cash and food crops will increase. In the opposite case, it will fall. Again, what will happen to the marketed surplus of food crops is not certain. The same considerations as before apply.

The Importance of Outside Alternatives

So far, we have considered farm households with no options to work outside of agriculture. The household survey carried out in 2001, however, reported two significant findings: (1) among the individuals who were in the labor force, those belonging to households whose heads were limited to farm employment were considerably poorer than when the heads worked in wage employment, household business, or other outside ventures and (2) almost 70 percent of all Timorese belonged to families with no or only small possibilities of obtaining income from outside agriculture.²¹⁶

This clearly points to the desirability of creating employment outside the agricultural sector. The conclusion is supported by later empirical evidence as well. The 2011 survey of coffee growers in Ermera stressed that ‘In the current rural economy, almost the only viable path out of extreme poverty is to find non-agricultural labour income’ and that ‘It appears as if the less poor households are able to achieve their higher incomes largely from labour income—roles such as teachers, drivers, security guards etc.’²¹⁷

In the short run, outside employment could for example take the form of public works during the off-peak season in agriculture. In the longer run, a private sector will develop. Smallholders in the Third World often derive their income from a variety of sources, and this serves to increase their standard of living. This finding can easily be incorporated into our model. In Figure 5.7, the possibility of earning a given outside wage equal to $\mathrm{w}/{\mathrm{p}}_{\mathrm{m}},$ when measured in terms of purchased goods, has been incorporated. The household keeps working in agriculture until the value of the marginal product of labor there falls to the level given by the constant outside wage. At this point $\left(\mathrm{B}\right),$ the income-leisure curve becomes a straight line. Thereafter, the remaining hours worked are spent in wage employment, up to point $\mathrm{C}.$ The income from this must be added to that from farm work. Hence, the total income line will lie outside $\mathrm{IJ}$ and outside the budget line $\mathrm{JK}.$ The incorporation of outside employment changes our conclusions.

Fig. 5.7

An increase in the relative price of cash crops

Let us begin with the case of increased prices of cash crops. When ${\mathrm{p}}_{\mathrm{c}}$ increases, as before, the income-leisure tradeoff rotates anti-clockwise, from $\mathrm{OA}$ to ${\mathrm{OA}}_1.$ The value of the marginal product increases in agriculture, and more hours will be worked there. Outside work will no longer begin at $\mathrm{B}$ (or $\mathrm{D}$) but at ${\mathrm{B}}_1$ (or ${\mathrm{D}}_1$). The total number of hours worked will, however, decrease. Provided that leisure is not an inferior good, points ${\mathrm{C}}_1$ and ${\mathrm{E}}_1$ lie to the right of $\mathrm{C}$ and $\mathrm{E},$ respectively, as long as some outside work takes place both before and after the price increase. The reason is that the substitution effect disappears. Leisure is no more expensive at the margin after the price increase than before, since the marginal valuation is given by the wage rate and not by the price of cash crops, and the wage rate does not change. All that remains is the income effect, and this makes for increased leisure. Income increases as well but as a result of the reallocation of work hours between wage labor and farm work. Should leisure on the other hand be inferior, ${\mathrm{C}}_1$ will be to the left of $\mathrm{C}$ and ${\mathrm{E}}_1$ to the left of $\mathrm{E}.$ Work will increase both inside and outside agriculture, at the expense of leisure.

Turning to the allocation of work hours between cash and food crop production, (5.8) may now be rewritten as

$$ \left({\mathrm{p}}_{\mathrm{f}}/{\mathrm{p}}_{\mathrm{m}}\right){{\mathrm{q}}_{\mathrm{f}}}^{\mathrm{L}}=\left({\mathrm{p}}_{\mathrm{c}}/{\mathrm{p}}_{\mathrm{m}}\right){{\mathrm{q}}_{\mathrm{c}}}^{\mathrm{L}}=\mathrm{w}/{\mathrm{p}}_{\mathrm{m}} $$

(5.15)

where the wage rate is given in terms of purchased goods. Also, the available labor time has to be distributed among leisure, crop production, and outside work. This changes (5.4) to

$$ \mathrm{L}={\mathrm{L}}_{\mathrm{e}}+{\mathrm{L}}_{\mathrm{f}}+{\mathrm{L}}_{\mathrm{c}}+{\mathrm{L}}_{\mathrm{w}} $$

(5.16)

Differentiating (5.15) when ${\mathrm{p}}_{\mathrm{c}}$ changes and the other two prices and the wage rate in terms of purchased goods $\left(\mathrm{w}/{\mathrm{p}}_{\mathrm{m}}\right)$ is held as constant (with ${\mathrm{p}}_{\mathrm{f}}={\mathrm{p}}_{\mathrm{c}}={\mathrm{p}}_{\mathrm{m}}=1$ at the outset) yields

$$ {{\mathrm{q}}_{\mathrm{f}}}^{\mathrm{L}\mathrm{L}}{\mathrm{dL}}_{\mathrm{f}}={{\mathrm{q}}_{\mathrm{c}}}^{\mathrm{L}\mathrm{L}}{\mathrm{dL}}_{\mathrm{c}}+{{\mathrm{q}}_{\mathrm{c}}}^{\mathrm{L}}{\mathrm{dp}}_{\mathrm{c}}=0 $$

(5.17)

or

$$ {\mathrm{dL}}_{\mathrm{c}}=-\left({{\mathrm{q}}_{\mathrm{c}}}^{\mathrm{L}}/{{\mathrm{q}}_{\mathrm{c}}}^{\mathrm{L}\mathrm{L}}\right){\mathrm{dp}}_{\mathrm{c}}>0 $$

(5.18)

and

$$ {\mathrm{dL}}_{\mathrm{f}}=0 $$

(5.19)

The output of cash crops increases while that of food remains constant. Here, price incentives work unequivocally, and since all cash crops are sold in the market, outside sales will increase as well. The marketed surplus of foodstuffs, on the other hand, will be reduced. Farm household income has increased, and both food prices and the price of purchased goods remain constant. Hence, only the income effect remains. The demand for both food and purchased goods among farm households will increase, but food output remains constant. The marketed surplus of food must therefore shrink.

The analysis is analogous for the case of increased food prices. There will be more leisure and larger food production, while cash crop production will remain constant. In this case, however, we don’t know with certainty what will happen to the marketed surplus of food. The price of the latter good increases and the substitution effect will be reintroduced. Possibly, on-farm consumption will be reduced, and more food will be sold outside.

The conclusions will change also when the price of purchased goods is reduced (but $\mathrm{w}/{\mathrm{p}}_{\mathrm{m}}$ remains constant). The income-leisure tradeoff again rotates anti-clockwise, more leisure is enjoyed, and less work takes place outside the farm and more inside. (5.12), in turn, becomes

$$ {{\mathrm{q}}_{\mathrm{f}}}^{\mathrm{L}\mathrm{L}}{\mathrm{dL}}_{\mathrm{f}}-{{\mathrm{q}}_{\mathrm{f}}}^{\mathrm{L}}{\mathrm{dp}}_{\mathrm{m}}={{\mathrm{q}}_{\mathrm{c}}}^{\mathrm{L}\mathrm{L}}{\mathrm{dL}}_{\mathrm{c}}-{{\mathrm{q}}_{\mathrm{c}}}^{\mathrm{L}}{\mathrm{dp}}_{\mathrm{m}}=0 $$

(5.20)

and

$$ {\mathrm{dL}}_{\mathrm{c}}=\left({{\mathrm{q}}_{\mathrm{c}}}^{\mathrm{L}}/{{\mathrm{q}}_{\mathrm{c}}}^{\mathrm{L}\mathrm{L}}\right){\mathrm{dp}}_{\mathrm{m}}>0 $$

(5.21)

$$ {\mathrm{dL}}_{\mathrm{f}}=\left({{\mathrm{q}}_{\mathrm{f}}}^{\mathrm{L}}/{{\mathrm{q}}_{\mathrm{f}}}^{\mathrm{L}\mathrm{L}}\right){\mathrm{dp}}_{\mathrm{m}}>0 $$

(5.22)

${\mathrm{dp}}_{\mathrm{m}}<0,$ so the production of both cash crops and food will increase. The farm households will increase their consumption of purchased goods and possibly, but not with certainty, that of food as well, since food becomes more expensive in terms of purchased goods. However, since the production of foodstuffs increases as well, the net result of the cheapening of purchased goods may be increased leisure, at the expense of outside work (provided that leisure is a superior good), increased on-farm consumption of food, and increased sales not only of cash crops but also of food.