GROWTH AND STAGNATION IN PAKISTAN AGRICULTURE

David Seckler 9

OVERVIEW

This report provides a preliminary analysis of the major factors behind agricultural production in Pakistan and especially attempts to account for the apparent stagnation of wheat and rice production in the 1980s It is hoped that this analysis will lead to much more detailed and disaggregated studies by members of the Economic Analysis Network and others in Pakistan

The central problem is shown in Figure 1 Wheat and ice production in Pakistan have been on something of a roller-coaster ride the past 15 years While production of both crops grew rapidly during the latter 1970s the rate of growth of wheat production slowed abruptly in the 1980s and rice productin hit a plateau Why this occurred is the major focus of attention

The answer it is believed is shown in the singularities noted in Figure 1 In particular the implementation of the New Agricultural Policy (NAP) of 1980 seems to correspond to the beginning of trouble Two major policy decisions appear to have adversely affected production

1) Investment in the expansion of irrigated area was reduced in favor of rehabilitation of existing irrigation systems and improved drainage Whatever the desirability or even necessity of rehabilitation and drainage may be it is expansion of irrigated area that is the basis for growing agricultural production in semiarid countries like Pakistan

2) Subsidies to fertilizers a d pesticides were reduced while simultaneously farm support prices for wheat and rice were ermitted to decrease in real terms The latter is of course another form of reduced subsidies to farmers and increased subsidies to consumers and export markets As a consequence Pakistani farmers have been put in a severe price squeeze so that it does not pay them to invest their resources in increased production

It is difficult to separate the individual effects of these two decisions as they happened more or less at the same time But this and other studies (Seckler nd Sampath Production and

Poverty in Indian Agriculture) indicate that in all but the most favored agro-climatic areas expansion of irrigated area rolls

out the carpet so to speak on which fertilizers HYVs labor and other inputs perform While biochemical technology can result in short-run gains in production it rather quickly tecomes weakened by diminishing returns to fertilizer and by saturation of crop areas under HYVs (which obviously cannot exceed 100) Thus in the long run agricultural production is

1

Figure 1

sHEAT AND RICE PRODUCTION (M

YV 72 -4

Si-41

NAP

- I -_ltI

I ~TarbeaA ---

- WHE47 + RICE

determined by basic (and unfortunately rare) breakthroughs in HYVs and especially by expansion of irrigated area

This year the GOP has substantially changed the price policyof the past few years by keeping fertilizer prices constant while increasing wheat support prices by 14 This policy changetogether with good weather has caused fertilizer use in the 1985-1986 wheat crop to increase by about the same percentage amount Consequently wheat production is expected to rise from the 115 MT level of last year to over 13 MT--again by about 14 (but not much more that the 125 MT level of 1981) While this is a gratifying result (which also corroborates the price squeeze hypothesis) it does not form a sufficient basis for long-term agricultural policy

Pakistan is experiencing severely diminishing returns to fertilizer use (as shown below in Table 2) This problem can only be solved by breakthroughs in HYVs expansion of irrigated area improved quality of irrigation and better farm management practices (which increase costs of production) While there is still om for crop price increases the tolerance of consumers and export markets is limited

Thus it appears that the major avenue for increased food production in Pakistan which has the potential for feeding a population growing at 31 pa is better management of existing irrigation systems and expansion of irrigation to new areas Here the word management of existing irrigation systems should be emphasized because simple rehabilitation of physical structures is not enough Both the physical and the managerial systems must be improved if these investments are to yield a favorable return In expansion of irrigated area tubewell development seems to offer the quickest and least expensive path However both these opportunities need very careful assessment because both have problems especially of a managerial nature

If as contended here it is necessary to expand irrigated area so that Pakistan can feed its rapidly growing population the question arises as to what will happen when the physicallimits of irrigation expansion are reached in the next two or three decades by which time population will have doubled Will biochemical technology and induced innovations be enough If there is some presently unforseeable breakthrough they could be but barring miracles it is most unlikely Without expansion of irrigated area what may be called the required rate of innovation to increase yields would be roughly the same as the rate of population growth over 3 pa This is far too much reasonably tc expect Pakistan must use the time left to reduce its long-run rate of population growth to about one-third the present rate This is the most important food policy decision facing Pakistan today

2

Figure 1

sHEAT AND RICE PRODUCTION (M

YV 72 -4

Si-41

NAP

- I -_ltI

I ~TarbeaA ---

- WHE47 + RICE

determined by basic (and unfortunately rare) breakthroughs in HYVs and especially by expansion of irrigated area

This year the GOP has substantially changed the price policyof the past few years by keeping fertilizer prices constant while increasing wheat support prices by 14 This policy changetogether with good weather has caused fertilizer use in the 1985-1986 wheat crop to increase by about the same percentage amount Consequently wheat production is expected to rise from the 115 MT level of last year to over 13 MT--again by about 14 (but not much more that the 125 MT level of 1981) While this is a gratifying result (which also corroborates the price squeeze hypothesis) it does not form a sufficient basis for long-term agricultural policy

Pakistan is experiencing severely diminishing returns to fertilizer use (as shown below in Table 2) This problem can only be solved by breakthroughs in HYVs expansion of irrigated area improved quality of irrigation and better farm management practices (which increase costs of production) While there is still om for crop price increases the tolerance of consumers and export markets is limited

Thus it appears that the major avenue for increased food production in Pakistan which has the potential for feeding a population growing at 31 pa is better management of existing irrigation systems and expansion of irrigation to new areas Here the word management of existing irrigation systems should be emphasized because simple rehabilitation of physical structures is not enough Both the physical and the managerial systems must be improved if these investments are to yield a favorable return In expansion of irrigated area tubewell development seems to offer the quickest and least expensive path However both these opportunities need very careful assessment because both have problems especially of a managerial nature

If as contended here it is necessary to expand irrigated area so that Pakistan can feed its rapidly growing population the question arises as to what will happen when the physicallimits of irrigation expansion are reached in the next two or three decades by which time population will have doubled Will biochemical technology and induced innovations be enough If there is some presently unforseeable breakthrough they could be but barring miracles it is most unlikely Without expansion of irrigated area what may be called the required rate of innovation to increase yields would be roughly the same as the rate of population growth over 3 pa This is far too much reasonably tc expect Pakistan must use the time left to reduce its long-run rate of population growth to about one-third the present rate This is the most important food policy decision facing Pakistan today

2

determined by basic (and unfortunately rare) breakthroughs in HYVs and especially by expansion of irrigated area

This year the GOP has substantially changed the price policyof the past few years by keeping fertilizer prices constant while increasing wheat support prices by 14 This policy changetogether with good weather has caused fertilizer use in the 1985-1986 wheat crop to increase by about the same percentage amount Consequently wheat production is expected to rise from the 115 MT level of last year to over 13 MT--again by about 14 (but not much more that the 125 MT level of 1981) While this is a gratifying result (which also corroborates the price squeeze hypothesis) it does not form a sufficient basis for long-term agricultural policy

Pakistan is experiencing severely diminishing returns to fertilizer use (as shown below in Table 2) This problem can only be solved by breakthroughs in HYVs expansion of irrigated area improved quality of irrigation and better farm management practices (which increase costs of production) While there is still om for crop price increases the tolerance of consumers and export markets is limited

Thus it appears that the major avenue for increased food production in Pakistan which has the potential for feeding a population growing at 31 pa is better management of existing irrigation systems and expansion of irrigation to new areas Here the word management of existing irrigation systems should be emphasized because simple rehabilitation of physical structures is not enough Both the physical and the managerial systems must be improved if these investments are to yield a favorable return In expansion of irrigated area tubewell development seems to offer the quickest and least expensive path However both these opportunities need very careful assessment because both have problems especially of a managerial nature

If as contended here it is necessary to expand irrigated area so that Pakistan can feed its rapidly growing population the question arises as to what will happen when the physicallimits of irrigation expansion are reached in the next two or three decades by which time population will have doubled Will biochemical technology and induced innovations be enough If there is some presently unforseeable breakthrough they could be but barring miracles it is most unlikely Without expansion of irrigated area what may be called the required rate of innovation to increase yields would be roughly the same as the rate of population growth over 3 pa This is far too much reasonably tc expect Pakistan must use the time left to reduce its long-run rate of population growth to about one-third the present rate This is the most important food policy decision facing Pakistan today

2

HYPOTHESES

Many hypotheses have been proposed to account for the stagnation in wheat and rice production during the 1980s

The Weather

The weather can Li a culprit but it usually does not influence trends as much as variations around trends Dr Zakir Hussain has analyzed rainfall data as part of this report(Appendix A) Variations in rainfall seem to have more effect on regional than on total production

Rising Water Tables and Salinity

This is a problem about which Pakistan has received considerable international attention but there is very little scientifically reliable data about it In any case it is most unlikely that these problems could cause the abrupt changes in the production trends of direct interest here Also since rice is more tolerant to water-logging and salinity than wheat these factors should affect wheat more than rice production Instead the decreasing trend for rice production is much more severe than for wheat

Weak Institutions in Research Extension and Marketing

While perhaps the research extension and marketing institutions are weak the fact is that these same institutions produced very good results in the 1970s and it is most unlikelythat they have abruptly deteriorated in the 1980s While it could be argued that these institutions were adequate for aneasy ride up the production function but are no longersufficient in a era of rapidly diminishing returns this argument seems rather post hoc

Agricultural Labor Shortages Resulting in Poor Farm Management Practices

Agricultural specialists generally agree that poor farm management practices are employed in Pakistan The question however is whether these practices have deteriorated in recent years because of labor shortages There has been substantial rural-urban migration in recent years and about one million workers have gone abroad out of a labor force of 24 million Pakistan like most developing countries needs much better data on rural employment wages and income Given this ]imitation the present guess is that labor shortages have had some but not a great deal of effect on the problem at hand

3

Reduced Growth of Irrigated Area and Poor Management of Existing Irrigation Systems

The potential effects of irrigation on crop production can be seen in Figure 1 in the dramatic increase of rice productionfollowing the Tarbela project By the same token as new irrigation development reached a plateau in the 1980s so too did rice and to a lesser extent wheat production With respect to the management of existing systems it is interesting to observe that the amount of water supplied per ha of irrigated land in rabi season has increased about 20 since the early 1970s while the amount of wheat produced per unit of water has remained about constant

It may also be noted that these labor-irrigation factors together may help explain why wheat yields in the Indian Punjab are about 3 tonnes per ha but only 2 tonnes per ha in the Pakistan Punjab Appendix B provides some interesting crossshycountry comparisons First there are 15 ha of agricultural land per agricultural worker in Pakistan compared to one ha in India Seventy percent of this land is irrigated in Pakistan compared to 23 in India Last it is generally agreed although more data is needed that the famous warabandi system of irrigation management is better managed in India than in Pakistan

A Farm Price Squeeze

As already noted the farm price squeeze and the lack of growth in irrigated area are the two major suspects in the search for Who done it

WHEAT PRODUCTION

This analysis focuses on wheat production A similar analysis of rice production should be done However an analysisof rice requires separation of the input-output relationships for basmati and HYVs and this task is far beyond the time available here

Many regression analyses were run in an attempt to account for the variations in wheat production from 1971 to 1986 The besc analysis turned out in the end to be surprisingly simple

First as shown in equation RI of Table 1 94 of the variation in wheat production can be accounted for by the rate of total consumption of fertilizer in the rabi season (about 80 of which is used in wheat) This equation is slightly improved in R2 which has the multiplicative interaction between fertilizer and the percent of the wheat area under HYVs as the independentvariable It should be noted that while many other variables and combinations were tried these were by far the best equations

4

TABLE I REGRESSION ANALYSES

EQUATION DEPENDENT CONSTANT PREDICTOR R2 F VALUE VARIABLE VARIABLES

I

---------------------------------------------------------------------------------

DO

RI LOG WHEAT PRODUCTION

0694188

LOS FERTILIZER

USE INRABI 0373371 094 1947 222

R2 LOG WHEAT PRODUCTION

619476E

LOG (XAREA INHYVS TIMES FERT INRABI

0286026 095 2324 203

R3 FERTILIZER USE INRABI 1OTAL IRRiGaTED APEA

-25 2329I 096 275 175

R4 CHANGE INFERT FEF CHANGE INRETURN HA WHEAT Ti(T-I 1T-1(T-2)

00055 0006 051 104 ---------- ----------------------------------------------------

ALL COEFFICIENTS SIGN FICANf AT DYER 99

200

qCL

The next question is this If fertilizer use (and HYVs)drives wheat production then what drives fertilizer use The answer shown in R3 is again very simple It is irrigation Ninety-five percent of the variance in fertilizer use in the rabi season is accounted for by the variance in total irrigated area

A view of the irrigation-fertilizer relationship is shown in Figure 2 It is seen that as irrigated area expanded from 1975 to 1981 primarily due to Tarbela fertilizer use grew rapidly However as growth in irrigated area slowed after that time fertilizer use also slowed

But there is also the fact that fertilizer use decreased considerably duriag the period 1980-1982 even though irrigated area was expanding This suggests that irrigation is not the only determinant of fertilizer use other factors must be considered as well The question of price ratios between wheat and fertilizer naturally springs to mind This is the subject of the next section

WHEAT-FERTILIZER PRICE RATIOS

Table 2 shows the basic price data Columns A and B show the support and wholesale market prices of wheat respectively Column C shows the price of urea Column D shows the result of dividing total production of wheat pa by 80 (the amount thought to be used in wheat) of the total urea use in the rabi season The result is the average product of urea in terms of wheat It is seen that the average productivity of urea fell from 356 kg of wheat per kg of urea in 1971 to only 167 kg in 1985 Thus there are strongly diminishing returns to urea use in wheat production

By dividing the price of urea by the price of wheat the amount of wheat required to purchase one kg of urea is obtained Column E shows that the wheat cost of urea in terms of support prices has been increasing over time In 1971 124 kg of wheat would purchase one kg of urea in 1984 160 kg of wheat was required The price situation improved greatly in 1986 because of the increase in wheat prices

Farmers decisions to purchase fertilizer are largely determined by the relationship between both the productivity of fertilizer in terms of the crop and relative prices between fertilizer and the crop Column F shows this relationship Here the productivity of urea in terms of wheat Column D is divided by the wheat cost of urea Column E This ratio gives the return in terms of wheat to the farmers investment in urea In 1971 an investment in one kg of urea would return 287 kg of wheat in 1986 even after the price rise the return would be only 130 kg of wheat Columns G and H show the same calculations based on the limited wholesale price data available

5

Figure 2

WH EAT lB

- - -15 - _e - -4+-

14 -I 13 -+ T 13

1I]-It -tI

4 I

Inl197l 1g2 193 194 1975 t197 g7 1 l90IM flg i HS wllBtgl I8 lgflD 19 G

o rEPT- 4 TO1L IPPI IDN1D

--------------------------------------------------------------------------------------------------

--------------------------------------------------------------------------------------------------

TABLE 2 WHEAT AND UREA PRICES

YEAR YEAR A B r D E F G H I J SUPPORT WHOLESALE UREA WHEAT SUPPORT SUPPORT WHOLESALE WHOLESALE UREA USE WHEAT PRICE PRICE PRICE PROD PER WHEAT RSRETURN WHEAT RSRETURN INWHEAT YEILD WHEAT WHEAT KG UREA COST UREA WHEATUREA COST UREA WHEATUREA RSIKG RSIKGi RSiKG KG46m (CA) (DIE) (CB) (DI) KGHA KGHA

---------------------------------------------------------------------------shy

1971 1771 046 C57 35 124 287 244 108 172 1972 37 124 306 21 11991)44 5 1 197 973 6 117C7 34 280 299 14 1974 1974 062 114 304 16A 161 328 1248 1975 1975 09 117 11 419 111 377 11B 355 51 1320 1976 1976 099 104 148 245 149 164 105 234 465 1422 1977 1977 099 111 136 254 137 185 112 226 451 1431 1978 1578 099 137 136 212 137 154 138 153 497 1316 1979 1979 121 135 128 192 106 182 111 173 620 1488 198) 1960 125 136 186 168 149 113 108 155 745 156P 1961 1981 145 155 192 179 132 15 107 168 733 1647 1982 1992 145 18 26 18c 163 n 12 145 1565 192 19t 16 193 25 174 166 10 12 145 77) 167i 144 IPE 16 21 25 148 160 092 125 118 803 142219p) 1985 175 256 156 146 107 827 1612 1986 19B6 2 256 167 128 130 843 1757

SUPOT PRICES FOP THE CROF YEAR IE 1986 1985-1986) PRICES INLAHORE CALANDEF YEAR

The consequence of decreasing returns to investment in urea is illustrated in Figure 3 Here urea application per ha (divided by 10) is related to the (support price) return variable (F of Table 2) and to the wheat price of urea in the preceding year This figure indicates that as long as the wheat return is about 16 kg or more farmers will invest in fertilizer but below that amount they are reluctant to do so With the low returns of the past five years the question is not why the rate of growth of fertilizer use--and consequently of wheat production--slowed but rather why it grew at all

Figure 4 indicates that a possible answer to this question may lie in wholesale rather than support wheat prices The wheat cost of urea in terms of wholesale prices has been more stable than the support price index Many regression analyses were made to incorporate these price effects into the fertilizer functions of Table 1 but the results were not wholly satisfactory Much more detailed analysis is needed on these relationships than was possible here A promising line of further research on this subject is shown in R4 of Table 1 Here the change in urea use per ha is related to the change in the (support price) return to urea in the preceding year An R2 value of 051 is quite good for change variables

Nevertheless it seems clear that Pakistani farmers are going to need more favorable price policies if they are going to invest in increased production Either crop prices must increase or urea (and other input prices) must decrease or both Which should it be

Table 3 provides some insight into this policy questionCurrent prices have been converted into real prices through the consumer price index (CPI) In real terms wheat and basmati prices have decreased over 20 since 1979 while IIRI-6 has only gone down in the past two years Real urea prices went upsubstantially in the early 1980s but are now down to roughly the 1979 level

The lower part of Table 3 carries the analysis further bydividing per capita GDP by the real prices Thus for example per capita GDP would purchase 531 kg of wheat in 1979 but 751 kgin 1985 a 42 increase as shown in the lower right-hand columns Only urea is at 1979 parity by this index While the ability to purchase much more food represents a substantial gainin consumers surplus over this period it seems that the processhas gone too far and some of the surplus should be reallocated back to producers as in the increased price support for wheat this year

CONCLUSIONS

It seems clear that fertilizer use and HYVs drive wheat production and that fertilizer use is in turn mainly driven bythe combination of expanded irrigated area and getting prices

6

Figure 3

NH E-T

21 ~

1 -I

-V7-4-4

i9-71 19-73 1975 1917 1BI 1B3 1

2- 4 SUP RET T-1 S SUP COST L

__- G _

9-7

Figure 4

AEAT

1

-4

- a

K

1971

WHLI-7 RET -

197- 3

E

4 -

C4-7 shy

HSL W

9-

ZC OT Tshy

191

-

1983 1985

UREAirHA

---------------------------------------------------------------------------------

--------------------------------------------------------------

TABLE 3PRICES AND INCOME

YEAR SUPPORTMARKET PRICES PER K6 REAL PRICES 1979 BASE WHEAT BASMATI IRRI-6 UREA CPI BASE WHEAT BASMATI IRRI-6 UREA

1979 121 161 081 128 296 1 121 161 081 128 1980 125 161 081 186 332 112 111 144 072 166 1981 145 188 097 192 375 127 114 148 077 152 1982 145 213 113 236 404 136 106 156 0P3 173 1983 16 22 123 256 418 141 113 156 087 181 19B4 16 225 128 256 462 156 103 144 082 164 1985 175 225 128 256 508 172 102 131 075 144 1986 2 233 133 256 559 1I9 106 123 070 136

YEAR RE4L PRICES INRELATION TO PE CAPITA GDP PERENT CHANGE OVE Fq73 (RUPEES 1959-1960 BASE EAL 6DFI PRICE PC GDF WHEAT BASMATI Ii-o UREA PC GDP WHEAT BASMATI IRRI-6 UREA

1975 642 531 39 73 502 100 10 100 100 100 19CC 666 598 464 922 402 104 113 116 116 0BF 1981 682 596 460 891 45) 106 112 115 112 090 1982 737 694 472 890 426 115 131 118 112 085 1983 736 650 472 845 406 115 122 118 107 081 1984 736 718 511 89- 449 115 135 128 113 089 1985 766 751 584 1027 514 119 142 147 130 102

SOURCE WORLD BANK 1985

right Subject to more detailed research than has been possiblehere several major policy issues emerge from this analysis

1) The development of new irrigated area seemed to have stopped in favor of rehabilitation of existing systems But simple physical rehabilitation only preserves the status quoante--in fact if not in theory--and thus the prime driver of Pakistans wheat production expansion of irrigated area has been stopped in its tracks

2) It is often thought that vertical production--iehigher yields--is the most economical path to agriculturaldevelopment This is a theory that has yet to be proven one about which the writer has serious reservations But even sovertical production does not come out of the air it comes out of inputs--of fertilizer HYVs and labor And the farmer must have the proper prices to encourage him to invest in these inputs It appears at least up to this year that agricultural policy in Pakistan has not been providing the proper price incentives and that crop prices should be increased Even substantial increases will leave consumers much better off than they were in 1979 relative to their income

Thus it appears to the writer that current fertilizer pricesshould be frozen for a year or two and that crop pricesespecially of wheat and basmati should be increased This mayprovide a short-run stimulus to crop production in Pakistan But in the long run the prospects for sustained growth without expansion of irrigated area and improved management of existingirrigation systems are dim Pakistan clearly needs to develop anew irrigation strategy based on an objective assessment of the experience of the past and careful quantitative assessments of the technical and managerial opportunities of the future

Finally unless Pakistan can control the growth of its population there is no foreseeable solution to the food problem

The author is Professor Department of Agricultural and Resource Economics Executive Director of the International School for Agricultural and Resource Development (ISARD) and Co-Director of the Colorado Institute of Irrigation Management (CIIM) at Colorado State University

This is Part A of a report to USAID Pakistan under a contract with Chemonics International Consulting Division

7

Appendix A WHEAT AND RICE RESPONSE TO RAINFALL IN THE INDUS BASIN (A TIME SERIES ANALYSIS) by Dr Zakir Hussain (USAID)

Wheat and rice are the two major food grain crops Of the total wheat

80 percent is cultivated in the irrigated areas and rice is almost

irrigated crop Both these crops experienced Green Revolution

technologies in the past two decades In the past two years

production of these crops is declining to an alarming extent Several

experts are attributirng this phenomena to unfavourable weather

condition In this note efforts have been made to help scan the

impact of rainfall in the critical months of growing season of rice

and wheat The ensuing discussion seperatly deals with the wheat and

rice

WHEAT PRODUCTIION RESPONSE TO RAINFALL

Wheat production and rainfall data was obtained from the Ministry of

food and agriculture for the past fifteen years (1971-1985)The

critical rainfall months December through January for wheat crop were

identified The same months were then deflated with the normal

rainfall of the these months ( normal rainfall is the average of

preceding twenty years) and eventualiy rainfall indicies were

calculated In similar fashion these indicies were identified as

independent variables for the provinces and overall equations The

equations are as follows

Y = a + bXi + 1 + e (1)

LnY= - + b LnXi + T + e (2) (i---ll)

Where

Y = production (000 tons) X = weather index Ln= Loge =

error termT= time trenda and b are parameters to be estimated

1

On Pakistan basis seven independent variables ( PunjabSind and

Pakistan linear and log linear rain indicies and trend) were

regressed against linear and log linear production of wheat Stepwis

regression was used to delineate the effect of each regressorThe

results are obtained in table 1 The analysis revealed that wheat

production has grown at 4 percent per annum and the coefficient of

trend variable was highly significant at 99 percent probability

level Of all the indicies Sind rainfall index turned out to be

significant perhaps it precipitates in the month of December in Sind

which is critical period for wheat crop tillering This will become

more clear from the foregoing analysisThere seems to be no problem

of auto-correlation (DW23) and explained variance is as high as 9

percent (R2 9)

The above two equations were fitted to Funjab data in log and

linear terms and stepwise regression was applied to help scan the

impact of rainfall in the critical months The results are again

reported in table 1 which shown that production grew at 3 percent

and January rain turned out to be critical month for this staple foot

crop Surprisingly enough the canal authorities close the canals in

the month of January which adversely effects the yield of wheat

There is no problem of auto-correlation (DW256) and R2 is fairly

high (93) In case of Sind provinc-e trend variable turned to be

highly significant and wheat production grows at annual rate of 5

percent Fuerthermore December rain has important bearing on the

productivity of this cropIt may nct bE out of place to mention that

wheat is sown 2-3 weeks ealier in Sind an compared to Funjab The

general fit of the equation is pretty good

Reduced Growth of Irrigated Area and Poor Management of Existing Irrigation Systems

The potential effects of irrigation on crop production can be seen in Figure 1 in the dramatic increase of rice productionfollowing the Tarbela project By the same token as new irrigation development reached a plateau in the 1980s so too did rice and to a lesser extent wheat production With respect to the management of existing systems it is interesting to observe that the amount of water supplied per ha of irrigated land in rabi season has increased about 20 since the early 1970s while the amount of wheat produced per unit of water has remained about constant

It may also be noted that these labor-irrigation factors together may help explain why wheat yields in the Indian Punjab are about 3 tonnes per ha but only 2 tonnes per ha in the Pakistan Punjab Appendix B provides some interesting crossshycountry comparisons First there are 15 ha of agricultural land per agricultural worker in Pakistan compared to one ha in India Seventy percent of this land is irrigated in Pakistan compared to 23 in India Last it is generally agreed although more data is needed that the famous warabandi system of irrigation management is better managed in India than in Pakistan

A Farm Price Squeeze

As already noted the farm price squeeze and the lack of growth in irrigated area are the two major suspects in the search for Who done it

WHEAT PRODUCTION

This analysis focuses on wheat production A similar analysis of rice production should be done However an analysisof rice requires separation of the input-output relationships for basmati and HYVs and this task is far beyond the time available here

Many regression analyses were run in an attempt to account for the variations in wheat production from 1971 to 1986 The besc analysis turned out in the end to be surprisingly simple

First as shown in equation RI of Table 1 94 of the variation in wheat production can be accounted for by the rate of total consumption of fertilizer in the rabi season (about 80 of which is used in wheat) This equation is slightly improved in R2 which has the multiplicative interaction between fertilizer and the percent of the wheat area under HYVs as the independentvariable It should be noted that while many other variables and combinations were tried these were by far the best equations

4

TABLE I REGRESSION ANALYSES

EQUATION DEPENDENT CONSTANT PREDICTOR R2 F VALUE VARIABLE VARIABLES

I

---------------------------------------------------------------------------------

DO

RI LOG WHEAT PRODUCTION

0694188

LOS FERTILIZER

USE INRABI 0373371 094 1947 222

R2 LOG WHEAT PRODUCTION

619476E

LOG (XAREA INHYVS TIMES FERT INRABI

0286026 095 2324 203

R3 FERTILIZER USE INRABI 1OTAL IRRiGaTED APEA

-25 2329I 096 275 175

R4 CHANGE INFERT FEF CHANGE INRETURN HA WHEAT Ti(T-I 1T-1(T-2)

00055 0006 051 104 ---------- ----------------------------------------------------

ALL COEFFICIENTS SIGN FICANf AT DYER 99

200

qCL

The next question is this If fertilizer use (and HYVs)drives wheat production then what drives fertilizer use The answer shown in R3 is again very simple It is irrigation Ninety-five percent of the variance in fertilizer use in the rabi season is accounted for by the variance in total irrigated area

A view of the irrigation-fertilizer relationship is shown in Figure 2 It is seen that as irrigated area expanded from 1975 to 1981 primarily due to Tarbela fertilizer use grew rapidly However as growth in irrigated area slowed after that time fertilizer use also slowed

But there is also the fact that fertilizer use decreased considerably duriag the period 1980-1982 even though irrigated area was expanding This suggests that irrigation is not the only determinant of fertilizer use other factors must be considered as well The question of price ratios between wheat and fertilizer naturally springs to mind This is the subject of the next section

WHEAT-FERTILIZER PRICE RATIOS

Table 2 shows the basic price data Columns A and B show the support and wholesale market prices of wheat respectively Column C shows the price of urea Column D shows the result of dividing total production of wheat pa by 80 (the amount thought to be used in wheat) of the total urea use in the rabi season The result is the average product of urea in terms of wheat It is seen that the average productivity of urea fell from 356 kg of wheat per kg of urea in 1971 to only 167 kg in 1985 Thus there are strongly diminishing returns to urea use in wheat production

By dividing the price of urea by the price of wheat the amount of wheat required to purchase one kg of urea is obtained Column E shows that the wheat cost of urea in terms of support prices has been increasing over time In 1971 124 kg of wheat would purchase one kg of urea in 1984 160 kg of wheat was required The price situation improved greatly in 1986 because of the increase in wheat prices

Farmers decisions to purchase fertilizer are largely determined by the relationship between both the productivity of fertilizer in terms of the crop and relative prices between fertilizer and the crop Column F shows this relationship Here the productivity of urea in terms of wheat Column D is divided by the wheat cost of urea Column E This ratio gives the return in terms of wheat to the farmers investment in urea In 1971 an investment in one kg of urea would return 287 kg of wheat in 1986 even after the price rise the return would be only 130 kg of wheat Columns G and H show the same calculations based on the limited wholesale price data available

5

Figure 2

WH EAT lB

- - -15 - _e - -4+-

14 -I 13 -+ T 13

1I]-It -tI

4 I

Inl197l 1g2 193 194 1975 t197 g7 1 l90IM flg i HS wllBtgl I8 lgflD 19 G

o rEPT- 4 TO1L IPPI IDN1D

--------------------------------------------------------------------------------------------------

--------------------------------------------------------------------------------------------------

TABLE 2 WHEAT AND UREA PRICES

YEAR YEAR A B r D E F G H I J SUPPORT WHOLESALE UREA WHEAT SUPPORT SUPPORT WHOLESALE WHOLESALE UREA USE WHEAT PRICE PRICE PRICE PROD PER WHEAT RSRETURN WHEAT RSRETURN INWHEAT YEILD WHEAT WHEAT KG UREA COST UREA WHEATUREA COST UREA WHEATUREA RSIKG RSIKGi RSiKG KG46m (CA) (DIE) (CB) (DI) KGHA KGHA

---------------------------------------------------------------------------shy

1971 1771 046 C57 35 124 287 244 108 172 1972 37 124 306 21 11991)44 5 1 197 973 6 117C7 34 280 299 14 1974 1974 062 114 304 16A 161 328 1248 1975 1975 09 117 11 419 111 377 11B 355 51 1320 1976 1976 099 104 148 245 149 164 105 234 465 1422 1977 1977 099 111 136 254 137 185 112 226 451 1431 1978 1578 099 137 136 212 137 154 138 153 497 1316 1979 1979 121 135 128 192 106 182 111 173 620 1488 198) 1960 125 136 186 168 149 113 108 155 745 156P 1961 1981 145 155 192 179 132 15 107 168 733 1647 1982 1992 145 18 26 18c 163 n 12 145 1565 192 19t 16 193 25 174 166 10 12 145 77) 167i 144 IPE 16 21 25 148 160 092 125 118 803 142219p) 1985 175 256 156 146 107 827 1612 1986 19B6 2 256 167 128 130 843 1757

SUPOT PRICES FOP THE CROF YEAR IE 1986 1985-1986) PRICES INLAHORE CALANDEF YEAR

The consequence of decreasing returns to investment in urea is illustrated in Figure 3 Here urea application per ha (divided by 10) is related to the (support price) return variable (F of Table 2) and to the wheat price of urea in the preceding year This figure indicates that as long as the wheat return is about 16 kg or more farmers will invest in fertilizer but below that amount they are reluctant to do so With the low returns of the past five years the question is not why the rate of growth of fertilizer use--and consequently of wheat production--slowed but rather why it grew at all

Figure 4 indicates that a possible answer to this question may lie in wholesale rather than support wheat prices The wheat cost of urea in terms of wholesale prices has been more stable than the support price index Many regression analyses were made to incorporate these price effects into the fertilizer functions of Table 1 but the results were not wholly satisfactory Much more detailed analysis is needed on these relationships than was possible here A promising line of further research on this subject is shown in R4 of Table 1 Here the change in urea use per ha is related to the change in the (support price) return to urea in the preceding year An R2 value of 051 is quite good for change variables

Nevertheless it seems clear that Pakistani farmers are going to need more favorable price policies if they are going to invest in increased production Either crop prices must increase or urea (and other input prices) must decrease or both Which should it be

Table 3 provides some insight into this policy questionCurrent prices have been converted into real prices through the consumer price index (CPI) In real terms wheat and basmati prices have decreased over 20 since 1979 while IIRI-6 has only gone down in the past two years Real urea prices went upsubstantially in the early 1980s but are now down to roughly the 1979 level

The lower part of Table 3 carries the analysis further bydividing per capita GDP by the real prices Thus for example per capita GDP would purchase 531 kg of wheat in 1979 but 751 kgin 1985 a 42 increase as shown in the lower right-hand columns Only urea is at 1979 parity by this index While the ability to purchase much more food represents a substantial gainin consumers surplus over this period it seems that the processhas gone too far and some of the surplus should be reallocated back to producers as in the increased price support for wheat this year

CONCLUSIONS

It seems clear that fertilizer use and HYVs drive wheat production and that fertilizer use is in turn mainly driven bythe combination of expanded irrigated area and getting prices

6

Figure 3

NH E-T

21 ~

1 -I

-V7-4-4

i9-71 19-73 1975 1917 1BI 1B3 1

2- 4 SUP RET T-1 S SUP COST L

__- G _

9-7

Figure 4

AEAT

1

-4

- a

K

1971

WHLI-7 RET -

197- 3

E

4 -

C4-7 shy

HSL W

9-

ZC OT Tshy

191

-

1983 1985

UREAirHA

---------------------------------------------------------------------------------

--------------------------------------------------------------

TABLE 3PRICES AND INCOME

YEAR SUPPORTMARKET PRICES PER K6 REAL PRICES 1979 BASE WHEAT BASMATI IRRI-6 UREA CPI BASE WHEAT BASMATI IRRI-6 UREA

1979 121 161 081 128 296 1 121 161 081 128 1980 125 161 081 186 332 112 111 144 072 166 1981 145 188 097 192 375 127 114 148 077 152 1982 145 213 113 236 404 136 106 156 0P3 173 1983 16 22 123 256 418 141 113 156 087 181 19B4 16 225 128 256 462 156 103 144 082 164 1985 175 225 128 256 508 172 102 131 075 144 1986 2 233 133 256 559 1I9 106 123 070 136

YEAR RE4L PRICES INRELATION TO PE CAPITA GDP PERENT CHANGE OVE Fq73 (RUPEES 1959-1960 BASE EAL 6DFI PRICE PC GDF WHEAT BASMATI Ii-o UREA PC GDP WHEAT BASMATI IRRI-6 UREA

1975 642 531 39 73 502 100 10 100 100 100 19CC 666 598 464 922 402 104 113 116 116 0BF 1981 682 596 460 891 45) 106 112 115 112 090 1982 737 694 472 890 426 115 131 118 112 085 1983 736 650 472 845 406 115 122 118 107 081 1984 736 718 511 89- 449 115 135 128 113 089 1985 766 751 584 1027 514 119 142 147 130 102

SOURCE WORLD BANK 1985

right Subject to more detailed research than has been possiblehere several major policy issues emerge from this analysis

1) The development of new irrigated area seemed to have stopped in favor of rehabilitation of existing systems But simple physical rehabilitation only preserves the status quoante--in fact if not in theory--and thus the prime driver of Pakistans wheat production expansion of irrigated area has been stopped in its tracks

2) It is often thought that vertical production--iehigher yields--is the most economical path to agriculturaldevelopment This is a theory that has yet to be proven one about which the writer has serious reservations But even sovertical production does not come out of the air it comes out of inputs--of fertilizer HYVs and labor And the farmer must have the proper prices to encourage him to invest in these inputs It appears at least up to this year that agricultural policy in Pakistan has not been providing the proper price incentives and that crop prices should be increased Even substantial increases will leave consumers much better off than they were in 1979 relative to their income

Thus it appears to the writer that current fertilizer pricesshould be frozen for a year or two and that crop pricesespecially of wheat and basmati should be increased This mayprovide a short-run stimulus to crop production in Pakistan But in the long run the prospects for sustained growth without expansion of irrigated area and improved management of existingirrigation systems are dim Pakistan clearly needs to develop anew irrigation strategy based on an objective assessment of the experience of the past and careful quantitative assessments of the technical and managerial opportunities of the future

Finally unless Pakistan can control the growth of its population there is no foreseeable solution to the food problem

The author is Professor Department of Agricultural and Resource Economics Executive Director of the International School for Agricultural and Resource Development (ISARD) and Co-Director of the Colorado Institute of Irrigation Management (CIIM) at Colorado State University

This is Part A of a report to USAID Pakistan under a contract with Chemonics International Consulting Division

7

Appendix A WHEAT AND RICE RESPONSE TO RAINFALL IN THE INDUS BASIN (A TIME SERIES ANALYSIS) by Dr Zakir Hussain (USAID)

Wheat and rice are the two major food grain crops Of the total wheat

80 percent is cultivated in the irrigated areas and rice is almost

irrigated crop Both these crops experienced Green Revolution

technologies in the past two decades In the past two years

production of these crops is declining to an alarming extent Several

experts are attributirng this phenomena to unfavourable weather

condition In this note efforts have been made to help scan the

impact of rainfall in the critical months of growing season of rice

and wheat The ensuing discussion seperatly deals with the wheat and

rice

WHEAT PRODUCTIION RESPONSE TO RAINFALL

Wheat production and rainfall data was obtained from the Ministry of

food and agriculture for the past fifteen years (1971-1985)The

critical rainfall months December through January for wheat crop were

identified The same months were then deflated with the normal

rainfall of the these months ( normal rainfall is the average of

preceding twenty years) and eventualiy rainfall indicies were

calculated In similar fashion these indicies were identified as

independent variables for the provinces and overall equations The

equations are as follows

Y = a + bXi + 1 + e (1)

LnY= - + b LnXi + T + e (2) (i---ll)

Where

Y = production (000 tons) X = weather index Ln= Loge =

error termT= time trenda and b are parameters to be estimated

1

On Pakistan basis seven independent variables ( PunjabSind and

Pakistan linear and log linear rain indicies and trend) were

regressed against linear and log linear production of wheat Stepwis

regression was used to delineate the effect of each regressorThe

results are obtained in table 1 The analysis revealed that wheat

production has grown at 4 percent per annum and the coefficient of

trend variable was highly significant at 99 percent probability

level Of all the indicies Sind rainfall index turned out to be

significant perhaps it precipitates in the month of December in Sind

which is critical period for wheat crop tillering This will become

more clear from the foregoing analysisThere seems to be no problem

of auto-correlation (DW23) and explained variance is as high as 9

percent (R2 9)

The above two equations were fitted to Funjab data in log and

linear terms and stepwise regression was applied to help scan the

impact of rainfall in the critical months The results are again

reported in table 1 which shown that production grew at 3 percent

and January rain turned out to be critical month for this staple foot

crop Surprisingly enough the canal authorities close the canals in

the month of January which adversely effects the yield of wheat

There is no problem of auto-correlation (DW256) and R2 is fairly

high (93) In case of Sind provinc-e trend variable turned to be

highly significant and wheat production grows at annual rate of 5

percent Fuerthermore December rain has important bearing on the

productivity of this cropIt may nct bE out of place to mention that

wheat is sown 2-3 weeks ealier in Sind an compared to Funjab The

general fit of the equation is pretty good

TABLE I REGRESSION ANALYSES

EQUATION DEPENDENT CONSTANT PREDICTOR R2 F VALUE VARIABLE VARIABLES

I

---------------------------------------------------------------------------------

DO

RI LOG WHEAT PRODUCTION

0694188

LOS FERTILIZER

USE INRABI 0373371 094 1947 222

R2 LOG WHEAT PRODUCTION

619476E

LOG (XAREA INHYVS TIMES FERT INRABI

0286026 095 2324 203

R3 FERTILIZER USE INRABI 1OTAL IRRiGaTED APEA

-25 2329I 096 275 175

R4 CHANGE INFERT FEF CHANGE INRETURN HA WHEAT Ti(T-I 1T-1(T-2)

00055 0006 051 104 ---------- ----------------------------------------------------

ALL COEFFICIENTS SIGN FICANf AT DYER 99

200

qCL

The next question is this If fertilizer use (and HYVs)drives wheat production then what drives fertilizer use The answer shown in R3 is again very simple It is irrigation Ninety-five percent of the variance in fertilizer use in the rabi season is accounted for by the variance in total irrigated area

A view of the irrigation-fertilizer relationship is shown in Figure 2 It is seen that as irrigated area expanded from 1975 to 1981 primarily due to Tarbela fertilizer use grew rapidly However as growth in irrigated area slowed after that time fertilizer use also slowed

But there is also the fact that fertilizer use decreased considerably duriag the period 1980-1982 even though irrigated area was expanding This suggests that irrigation is not the only determinant of fertilizer use other factors must be considered as well The question of price ratios between wheat and fertilizer naturally springs to mind This is the subject of the next section

WHEAT-FERTILIZER PRICE RATIOS

Table 2 shows the basic price data Columns A and B show the support and wholesale market prices of wheat respectively Column C shows the price of urea Column D shows the result of dividing total production of wheat pa by 80 (the amount thought to be used in wheat) of the total urea use in the rabi season The result is the average product of urea in terms of wheat It is seen that the average productivity of urea fell from 356 kg of wheat per kg of urea in 1971 to only 167 kg in 1985 Thus there are strongly diminishing returns to urea use in wheat production

By dividing the price of urea by the price of wheat the amount of wheat required to purchase one kg of urea is obtained Column E shows that the wheat cost of urea in terms of support prices has been increasing over time In 1971 124 kg of wheat would purchase one kg of urea in 1984 160 kg of wheat was required The price situation improved greatly in 1986 because of the increase in wheat prices

Farmers decisions to purchase fertilizer are largely determined by the relationship between both the productivity of fertilizer in terms of the crop and relative prices between fertilizer and the crop Column F shows this relationship Here the productivity of urea in terms of wheat Column D is divided by the wheat cost of urea Column E This ratio gives the return in terms of wheat to the farmers investment in urea In 1971 an investment in one kg of urea would return 287 kg of wheat in 1986 even after the price rise the return would be only 130 kg of wheat Columns G and H show the same calculations based on the limited wholesale price data available

5

Figure 2

WH EAT lB

- - -15 - _e - -4+-

14 -I 13 -+ T 13

1I]-It -tI

4 I

Inl197l 1g2 193 194 1975 t197 g7 1 l90IM flg i HS wllBtgl I8 lgflD 19 G

o rEPT- 4 TO1L IPPI IDN1D

--------------------------------------------------------------------------------------------------

--------------------------------------------------------------------------------------------------

TABLE 2 WHEAT AND UREA PRICES

YEAR YEAR A B r D E F G H I J SUPPORT WHOLESALE UREA WHEAT SUPPORT SUPPORT WHOLESALE WHOLESALE UREA USE WHEAT PRICE PRICE PRICE PROD PER WHEAT RSRETURN WHEAT RSRETURN INWHEAT YEILD WHEAT WHEAT KG UREA COST UREA WHEATUREA COST UREA WHEATUREA RSIKG RSIKGi RSiKG KG46m (CA) (DIE) (CB) (DI) KGHA KGHA

---------------------------------------------------------------------------shy

1971 1771 046 C57 35 124 287 244 108 172 1972 37 124 306 21 11991)44 5 1 197 973 6 117C7 34 280 299 14 1974 1974 062 114 304 16A 161 328 1248 1975 1975 09 117 11 419 111 377 11B 355 51 1320 1976 1976 099 104 148 245 149 164 105 234 465 1422 1977 1977 099 111 136 254 137 185 112 226 451 1431 1978 1578 099 137 136 212 137 154 138 153 497 1316 1979 1979 121 135 128 192 106 182 111 173 620 1488 198) 1960 125 136 186 168 149 113 108 155 745 156P 1961 1981 145 155 192 179 132 15 107 168 733 1647 1982 1992 145 18 26 18c 163 n 12 145 1565 192 19t 16 193 25 174 166 10 12 145 77) 167i 144 IPE 16 21 25 148 160 092 125 118 803 142219p) 1985 175 256 156 146 107 827 1612 1986 19B6 2 256 167 128 130 843 1757

SUPOT PRICES FOP THE CROF YEAR IE 1986 1985-1986) PRICES INLAHORE CALANDEF YEAR

The consequence of decreasing returns to investment in urea is illustrated in Figure 3 Here urea application per ha (divided by 10) is related to the (support price) return variable (F of Table 2) and to the wheat price of urea in the preceding year This figure indicates that as long as the wheat return is about 16 kg or more farmers will invest in fertilizer but below that amount they are reluctant to do so With the low returns of the past five years the question is not why the rate of growth of fertilizer use--and consequently of wheat production--slowed but rather why it grew at all

Figure 4 indicates that a possible answer to this question may lie in wholesale rather than support wheat prices The wheat cost of urea in terms of wholesale prices has been more stable than the support price index Many regression analyses were made to incorporate these price effects into the fertilizer functions of Table 1 but the results were not wholly satisfactory Much more detailed analysis is needed on these relationships than was possible here A promising line of further research on this subject is shown in R4 of Table 1 Here the change in urea use per ha is related to the change in the (support price) return to urea in the preceding year An R2 value of 051 is quite good for change variables

Nevertheless it seems clear that Pakistani farmers are going to need more favorable price policies if they are going to invest in increased production Either crop prices must increase or urea (and other input prices) must decrease or both Which should it be

Table 3 provides some insight into this policy questionCurrent prices have been converted into real prices through the consumer price index (CPI) In real terms wheat and basmati prices have decreased over 20 since 1979 while IIRI-6 has only gone down in the past two years Real urea prices went upsubstantially in the early 1980s but are now down to roughly the 1979 level

The lower part of Table 3 carries the analysis further bydividing per capita GDP by the real prices Thus for example per capita GDP would purchase 531 kg of wheat in 1979 but 751 kgin 1985 a 42 increase as shown in the lower right-hand columns Only urea is at 1979 parity by this index While the ability to purchase much more food represents a substantial gainin consumers surplus over this period it seems that the processhas gone too far and some of the surplus should be reallocated back to producers as in the increased price support for wheat this year

CONCLUSIONS

It seems clear that fertilizer use and HYVs drive wheat production and that fertilizer use is in turn mainly driven bythe combination of expanded irrigated area and getting prices

6

Figure 3

NH E-T

21 ~

1 -I

-V7-4-4

i9-71 19-73 1975 1917 1BI 1B3 1

2- 4 SUP RET T-1 S SUP COST L

__- G _

9-7

Figure 4

AEAT

1

-4

- a

K

1971

WHLI-7 RET -

197- 3

E

4 -

C4-7 shy

HSL W

9-

ZC OT Tshy

191

-

1983 1985

UREAirHA

---------------------------------------------------------------------------------

--------------------------------------------------------------

TABLE 3PRICES AND INCOME

YEAR SUPPORTMARKET PRICES PER K6 REAL PRICES 1979 BASE WHEAT BASMATI IRRI-6 UREA CPI BASE WHEAT BASMATI IRRI-6 UREA

1979 121 161 081 128 296 1 121 161 081 128 1980 125 161 081 186 332 112 111 144 072 166 1981 145 188 097 192 375 127 114 148 077 152 1982 145 213 113 236 404 136 106 156 0P3 173 1983 16 22 123 256 418 141 113 156 087 181 19B4 16 225 128 256 462 156 103 144 082 164 1985 175 225 128 256 508 172 102 131 075 144 1986 2 233 133 256 559 1I9 106 123 070 136

YEAR RE4L PRICES INRELATION TO PE CAPITA GDP PERENT CHANGE OVE Fq73 (RUPEES 1959-1960 BASE EAL 6DFI PRICE PC GDF WHEAT BASMATI Ii-o UREA PC GDP WHEAT BASMATI IRRI-6 UREA

1975 642 531 39 73 502 100 10 100 100 100 19CC 666 598 464 922 402 104 113 116 116 0BF 1981 682 596 460 891 45) 106 112 115 112 090 1982 737 694 472 890 426 115 131 118 112 085 1983 736 650 472 845 406 115 122 118 107 081 1984 736 718 511 89- 449 115 135 128 113 089 1985 766 751 584 1027 514 119 142 147 130 102

SOURCE WORLD BANK 1985

right Subject to more detailed research than has been possiblehere several major policy issues emerge from this analysis

1) The development of new irrigated area seemed to have stopped in favor of rehabilitation of existing systems But simple physical rehabilitation only preserves the status quoante--in fact if not in theory--and thus the prime driver of Pakistans wheat production expansion of irrigated area has been stopped in its tracks

2) It is often thought that vertical production--iehigher yields--is the most economical path to agriculturaldevelopment This is a theory that has yet to be proven one about which the writer has serious reservations But even sovertical production does not come out of the air it comes out of inputs--of fertilizer HYVs and labor And the farmer must have the proper prices to encourage him to invest in these inputs It appears at least up to this year that agricultural policy in Pakistan has not been providing the proper price incentives and that crop prices should be increased Even substantial increases will leave consumers much better off than they were in 1979 relative to their income

Thus it appears to the writer that current fertilizer pricesshould be frozen for a year or two and that crop pricesespecially of wheat and basmati should be increased This mayprovide a short-run stimulus to crop production in Pakistan But in the long run the prospects for sustained growth without expansion of irrigated area and improved management of existingirrigation systems are dim Pakistan clearly needs to develop anew irrigation strategy based on an objective assessment of the experience of the past and careful quantitative assessments of the technical and managerial opportunities of the future

Finally unless Pakistan can control the growth of its population there is no foreseeable solution to the food problem

The author is Professor Department of Agricultural and Resource Economics Executive Director of the International School for Agricultural and Resource Development (ISARD) and Co-Director of the Colorado Institute of Irrigation Management (CIIM) at Colorado State University

This is Part A of a report to USAID Pakistan under a contract with Chemonics International Consulting Division

7

Appendix A WHEAT AND RICE RESPONSE TO RAINFALL IN THE INDUS BASIN (A TIME SERIES ANALYSIS) by Dr Zakir Hussain (USAID)

Wheat and rice are the two major food grain crops Of the total wheat

80 percent is cultivated in the irrigated areas and rice is almost

irrigated crop Both these crops experienced Green Revolution

technologies in the past two decades In the past two years

production of these crops is declining to an alarming extent Several

experts are attributirng this phenomena to unfavourable weather

condition In this note efforts have been made to help scan the

impact of rainfall in the critical months of growing season of rice

and wheat The ensuing discussion seperatly deals with the wheat and

rice

WHEAT PRODUCTIION RESPONSE TO RAINFALL

Wheat production and rainfall data was obtained from the Ministry of

food and agriculture for the past fifteen years (1971-1985)The

critical rainfall months December through January for wheat crop were

identified The same months were then deflated with the normal

rainfall of the these months ( normal rainfall is the average of

preceding twenty years) and eventualiy rainfall indicies were

calculated In similar fashion these indicies were identified as

independent variables for the provinces and overall equations The

equations are as follows

Y = a + bXi + 1 + e (1)

LnY= - + b LnXi + T + e (2) (i---ll)

Where

Y = production (000 tons) X = weather index Ln= Loge =

error termT= time trenda and b are parameters to be estimated

1

On Pakistan basis seven independent variables ( PunjabSind and

Pakistan linear and log linear rain indicies and trend) were

regressed against linear and log linear production of wheat Stepwis

regression was used to delineate the effect of each regressorThe

results are obtained in table 1 The analysis revealed that wheat

production has grown at 4 percent per annum and the coefficient of

trend variable was highly significant at 99 percent probability

level Of all the indicies Sind rainfall index turned out to be

significant perhaps it precipitates in the month of December in Sind

which is critical period for wheat crop tillering This will become

more clear from the foregoing analysisThere seems to be no problem

of auto-correlation (DW23) and explained variance is as high as 9

percent (R2 9)

The above two equations were fitted to Funjab data in log and

linear terms and stepwise regression was applied to help scan the

impact of rainfall in the critical months The results are again

reported in table 1 which shown that production grew at 3 percent

and January rain turned out to be critical month for this staple foot

crop Surprisingly enough the canal authorities close the canals in

the month of January which adversely effects the yield of wheat

There is no problem of auto-correlation (DW256) and R2 is fairly

high (93) In case of Sind provinc-e trend variable turned to be

highly significant and wheat production grows at annual rate of 5

percent Fuerthermore December rain has important bearing on the

productivity of this cropIt may nct bE out of place to mention that

wheat is sown 2-3 weeks ealier in Sind an compared to Funjab The

general fit of the equation is pretty good

The next question is this If fertilizer use (and HYVs)drives wheat production then what drives fertilizer use The answer shown in R3 is again very simple It is irrigation Ninety-five percent of the variance in fertilizer use in the rabi season is accounted for by the variance in total irrigated area

A view of the irrigation-fertilizer relationship is shown in Figure 2 It is seen that as irrigated area expanded from 1975 to 1981 primarily due to Tarbela fertilizer use grew rapidly However as growth in irrigated area slowed after that time fertilizer use also slowed

But there is also the fact that fertilizer use decreased considerably duriag the period 1980-1982 even though irrigated area was expanding This suggests that irrigation is not the only determinant of fertilizer use other factors must be considered as well The question of price ratios between wheat and fertilizer naturally springs to mind This is the subject of the next section

WHEAT-FERTILIZER PRICE RATIOS

Table 2 shows the basic price data Columns A and B show the support and wholesale market prices of wheat respectively Column C shows the price of urea Column D shows the result of dividing total production of wheat pa by 80 (the amount thought to be used in wheat) of the total urea use in the rabi season The result is the average product of urea in terms of wheat It is seen that the average productivity of urea fell from 356 kg of wheat per kg of urea in 1971 to only 167 kg in 1985 Thus there are strongly diminishing returns to urea use in wheat production

By dividing the price of urea by the price of wheat the amount of wheat required to purchase one kg of urea is obtained Column E shows that the wheat cost of urea in terms of support prices has been increasing over time In 1971 124 kg of wheat would purchase one kg of urea in 1984 160 kg of wheat was required The price situation improved greatly in 1986 because of the increase in wheat prices

Farmers decisions to purchase fertilizer are largely determined by the relationship between both the productivity of fertilizer in terms of the crop and relative prices between fertilizer and the crop Column F shows this relationship Here the productivity of urea in terms of wheat Column D is divided by the wheat cost of urea Column E This ratio gives the return in terms of wheat to the farmers investment in urea In 1971 an investment in one kg of urea would return 287 kg of wheat in 1986 even after the price rise the return would be only 130 kg of wheat Columns G and H show the same calculations based on the limited wholesale price data available

5

Figure 2

WH EAT lB

- - -15 - _e - -4+-

14 -I 13 -+ T 13

1I]-It -tI

4 I

Inl197l 1g2 193 194 1975 t197 g7 1 l90IM flg i HS wllBtgl I8 lgflD 19 G

o rEPT- 4 TO1L IPPI IDN1D

--------------------------------------------------------------------------------------------------

--------------------------------------------------------------------------------------------------

TABLE 2 WHEAT AND UREA PRICES

YEAR YEAR A B r D E F G H I J SUPPORT WHOLESALE UREA WHEAT SUPPORT SUPPORT WHOLESALE WHOLESALE UREA USE WHEAT PRICE PRICE PRICE PROD PER WHEAT RSRETURN WHEAT RSRETURN INWHEAT YEILD WHEAT WHEAT KG UREA COST UREA WHEATUREA COST UREA WHEATUREA RSIKG RSIKGi RSiKG KG46m (CA) (DIE) (CB) (DI) KGHA KGHA

---------------------------------------------------------------------------shy

1971 1771 046 C57 35 124 287 244 108 172 1972 37 124 306 21 11991)44 5 1 197 973 6 117C7 34 280 299 14 1974 1974 062 114 304 16A 161 328 1248 1975 1975 09 117 11 419 111 377 11B 355 51 1320 1976 1976 099 104 148 245 149 164 105 234 465 1422 1977 1977 099 111 136 254 137 185 112 226 451 1431 1978 1578 099 137 136 212 137 154 138 153 497 1316 1979 1979 121 135 128 192 106 182 111 173 620 1488 198) 1960 125 136 186 168 149 113 108 155 745 156P 1961 1981 145 155 192 179 132 15 107 168 733 1647 1982 1992 145 18 26 18c 163 n 12 145 1565 192 19t 16 193 25 174 166 10 12 145 77) 167i 144 IPE 16 21 25 148 160 092 125 118 803 142219p) 1985 175 256 156 146 107 827 1612 1986 19B6 2 256 167 128 130 843 1757

SUPOT PRICES FOP THE CROF YEAR IE 1986 1985-1986) PRICES INLAHORE CALANDEF YEAR

The consequence of decreasing returns to investment in urea is illustrated in Figure 3 Here urea application per ha (divided by 10) is related to the (support price) return variable (F of Table 2) and to the wheat price of urea in the preceding year This figure indicates that as long as the wheat return is about 16 kg or more farmers will invest in fertilizer but below that amount they are reluctant to do so With the low returns of the past five years the question is not why the rate of growth of fertilizer use--and consequently of wheat production--slowed but rather why it grew at all

Figure 4 indicates that a possible answer to this question may lie in wholesale rather than support wheat prices The wheat cost of urea in terms of wholesale prices has been more stable than the support price index Many regression analyses were made to incorporate these price effects into the fertilizer functions of Table 1 but the results were not wholly satisfactory Much more detailed analysis is needed on these relationships than was possible here A promising line of further research on this subject is shown in R4 of Table 1 Here the change in urea use per ha is related to the change in the (support price) return to urea in the preceding year An R2 value of 051 is quite good for change variables

Nevertheless it seems clear that Pakistani farmers are going to need more favorable price policies if they are going to invest in increased production Either crop prices must increase or urea (and other input prices) must decrease or both Which should it be

Table 3 provides some insight into this policy questionCurrent prices have been converted into real prices through the consumer price index (CPI) In real terms wheat and basmati prices have decreased over 20 since 1979 while IIRI-6 has only gone down in the past two years Real urea prices went upsubstantially in the early 1980s but are now down to roughly the 1979 level

The lower part of Table 3 carries the analysis further bydividing per capita GDP by the real prices Thus for example per capita GDP would purchase 531 kg of wheat in 1979 but 751 kgin 1985 a 42 increase as shown in the lower right-hand columns Only urea is at 1979 parity by this index While the ability to purchase much more food represents a substantial gainin consumers surplus over this period it seems that the processhas gone too far and some of the surplus should be reallocated back to producers as in the increased price support for wheat this year

CONCLUSIONS

It seems clear that fertilizer use and HYVs drive wheat production and that fertilizer use is in turn mainly driven bythe combination of expanded irrigated area and getting prices

6

Figure 3

NH E-T

21 ~

1 -I

-V7-4-4

i9-71 19-73 1975 1917 1BI 1B3 1

2- 4 SUP RET T-1 S SUP COST L

__- G _

9-7

Figure 4

AEAT

1

-4

- a

K

1971

WHLI-7 RET -

197- 3

E

4 -

C4-7 shy

HSL W

9-

ZC OT Tshy

191

-

1983 1985

UREAirHA

---------------------------------------------------------------------------------

--------------------------------------------------------------

TABLE 3PRICES AND INCOME

YEAR SUPPORTMARKET PRICES PER K6 REAL PRICES 1979 BASE WHEAT BASMATI IRRI-6 UREA CPI BASE WHEAT BASMATI IRRI-6 UREA

1979 121 161 081 128 296 1 121 161 081 128 1980 125 161 081 186 332 112 111 144 072 166 1981 145 188 097 192 375 127 114 148 077 152 1982 145 213 113 236 404 136 106 156 0P3 173 1983 16 22 123 256 418 141 113 156 087 181 19B4 16 225 128 256 462 156 103 144 082 164 1985 175 225 128 256 508 172 102 131 075 144 1986 2 233 133 256 559 1I9 106 123 070 136

YEAR RE4L PRICES INRELATION TO PE CAPITA GDP PERENT CHANGE OVE Fq73 (RUPEES 1959-1960 BASE EAL 6DFI PRICE PC GDF WHEAT BASMATI Ii-o UREA PC GDP WHEAT BASMATI IRRI-6 UREA

1975 642 531 39 73 502 100 10 100 100 100 19CC 666 598 464 922 402 104 113 116 116 0BF 1981 682 596 460 891 45) 106 112 115 112 090 1982 737 694 472 890 426 115 131 118 112 085 1983 736 650 472 845 406 115 122 118 107 081 1984 736 718 511 89- 449 115 135 128 113 089 1985 766 751 584 1027 514 119 142 147 130 102

SOURCE WORLD BANK 1985

right Subject to more detailed research than has been possiblehere several major policy issues emerge from this analysis

1) The development of new irrigated area seemed to have stopped in favor of rehabilitation of existing systems But simple physical rehabilitation only preserves the status quoante--in fact if not in theory--and thus the prime driver of Pakistans wheat production expansion of irrigated area has been stopped in its tracks

2) It is often thought that vertical production--iehigher yields--is the most economical path to agriculturaldevelopment This is a theory that has yet to be proven one about which the writer has serious reservations But even sovertical production does not come out of the air it comes out of inputs--of fertilizer HYVs and labor And the farmer must have the proper prices to encourage him to invest in these inputs It appears at least up to this year that agricultural policy in Pakistan has not been providing the proper price incentives and that crop prices should be increased Even substantial increases will leave consumers much better off than they were in 1979 relative to their income

Thus it appears to the writer that current fertilizer pricesshould be frozen for a year or two and that crop pricesespecially of wheat and basmati should be increased This mayprovide a short-run stimulus to crop production in Pakistan But in the long run the prospects for sustained growth without expansion of irrigated area and improved management of existingirrigation systems are dim Pakistan clearly needs to develop anew irrigation strategy based on an objective assessment of the experience of the past and careful quantitative assessments of the technical and managerial opportunities of the future

Finally unless Pakistan can control the growth of its population there is no foreseeable solution to the food problem

The author is Professor Department of Agricultural and Resource Economics Executive Director of the International School for Agricultural and Resource Development (ISARD) and Co-Director of the Colorado Institute of Irrigation Management (CIIM) at Colorado State University

This is Part A of a report to USAID Pakistan under a contract with Chemonics International Consulting Division

7

Appendix A WHEAT AND RICE RESPONSE TO RAINFALL IN THE INDUS BASIN (A TIME SERIES ANALYSIS) by Dr Zakir Hussain (USAID)

Wheat and rice are the two major food grain crops Of the total wheat

80 percent is cultivated in the irrigated areas and rice is almost

irrigated crop Both these crops experienced Green Revolution

technologies in the past two decades In the past two years

production of these crops is declining to an alarming extent Several

experts are attributirng this phenomena to unfavourable weather

condition In this note efforts have been made to help scan the

impact of rainfall in the critical months of growing season of rice

and wheat The ensuing discussion seperatly deals with the wheat and

rice

WHEAT PRODUCTIION RESPONSE TO RAINFALL

Wheat production and rainfall data was obtained from the Ministry of

food and agriculture for the past fifteen years (1971-1985)The

critical rainfall months December through January for wheat crop were

identified The same months were then deflated with the normal

rainfall of the these months ( normal rainfall is the average of

preceding twenty years) and eventualiy rainfall indicies were

calculated In similar fashion these indicies were identified as

independent variables for the provinces and overall equations The

equations are as follows

Y = a + bXi + 1 + e (1)

LnY= - + b LnXi + T + e (2) (i---ll)

Where

Y = production (000 tons) X = weather index Ln= Loge =

error termT= time trenda and b are parameters to be estimated

1

On Pakistan basis seven independent variables ( PunjabSind and

Pakistan linear and log linear rain indicies and trend) were

regressed against linear and log linear production of wheat Stepwis

regression was used to delineate the effect of each regressorThe

results are obtained in table 1 The analysis revealed that wheat

production has grown at 4 percent per annum and the coefficient of

trend variable was highly significant at 99 percent probability

level Of all the indicies Sind rainfall index turned out to be

significant perhaps it precipitates in the month of December in Sind

which is critical period for wheat crop tillering This will become

more clear from the foregoing analysisThere seems to be no problem

of auto-correlation (DW23) and explained variance is as high as 9

percent (R2 9)

The above two equations were fitted to Funjab data in log and

linear terms and stepwise regression was applied to help scan the

impact of rainfall in the critical months The results are again

reported in table 1 which shown that production grew at 3 percent

and January rain turned out to be critical month for this staple foot

crop Surprisingly enough the canal authorities close the canals in

the month of January which adversely effects the yield of wheat

There is no problem of auto-correlation (DW256) and R2 is fairly

high (93) In case of Sind provinc-e trend variable turned to be

highly significant and wheat production grows at annual rate of 5

percent Fuerthermore December rain has important bearing on the

productivity of this cropIt may nct bE out of place to mention that

wheat is sown 2-3 weeks ealier in Sind an compared to Funjab The

general fit of the equation is pretty good

Figure 2

WH EAT lB

- - -15 - _e - -4+-

14 -I 13 -+ T 13

1I]-It -tI

4 I

Inl197l 1g2 193 194 1975 t197 g7 1 l90IM flg i HS wllBtgl I8 lgflD 19 G

o rEPT- 4 TO1L IPPI IDN1D

--------------------------------------------------------------------------------------------------

--------------------------------------------------------------------------------------------------

TABLE 2 WHEAT AND UREA PRICES

YEAR YEAR A B r D E F G H I J SUPPORT WHOLESALE UREA WHEAT SUPPORT SUPPORT WHOLESALE WHOLESALE UREA USE WHEAT PRICE PRICE PRICE PROD PER WHEAT RSRETURN WHEAT RSRETURN INWHEAT YEILD WHEAT WHEAT KG UREA COST UREA WHEATUREA COST UREA WHEATUREA RSIKG RSIKGi RSiKG KG46m (CA) (DIE) (CB) (DI) KGHA KGHA

---------------------------------------------------------------------------shy

1971 1771 046 C57 35 124 287 244 108 172 1972 37 124 306 21 11991)44 5 1 197 973 6 117C7 34 280 299 14 1974 1974 062 114 304 16A 161 328 1248 1975 1975 09 117 11 419 111 377 11B 355 51 1320 1976 1976 099 104 148 245 149 164 105 234 465 1422 1977 1977 099 111 136 254 137 185 112 226 451 1431 1978 1578 099 137 136 212 137 154 138 153 497 1316 1979 1979 121 135 128 192 106 182 111 173 620 1488 198) 1960 125 136 186 168 149 113 108 155 745 156P 1961 1981 145 155 192 179 132 15 107 168 733 1647 1982 1992 145 18 26 18c 163 n 12 145 1565 192 19t 16 193 25 174 166 10 12 145 77) 167i 144 IPE 16 21 25 148 160 092 125 118 803 142219p) 1985 175 256 156 146 107 827 1612 1986 19B6 2 256 167 128 130 843 1757

SUPOT PRICES FOP THE CROF YEAR IE 1986 1985-1986) PRICES INLAHORE CALANDEF YEAR

The consequence of decreasing returns to investment in urea is illustrated in Figure 3 Here urea application per ha (divided by 10) is related to the (support price) return variable (F of Table 2) and to the wheat price of urea in the preceding year This figure indicates that as long as the wheat return is about 16 kg or more farmers will invest in fertilizer but below that amount they are reluctant to do so With the low returns of the past five years the question is not why the rate of growth of fertilizer use--and consequently of wheat production--slowed but rather why it grew at all

Figure 4 indicates that a possible answer to this question may lie in wholesale rather than support wheat prices The wheat cost of urea in terms of wholesale prices has been more stable than the support price index Many regression analyses were made to incorporate these price effects into the fertilizer functions of Table 1 but the results were not wholly satisfactory Much more detailed analysis is needed on these relationships than was possible here A promising line of further research on this subject is shown in R4 of Table 1 Here the change in urea use per ha is related to the change in the (support price) return to urea in the preceding year An R2 value of 051 is quite good for change variables

Nevertheless it seems clear that Pakistani farmers are going to need more favorable price policies if they are going to invest in increased production Either crop prices must increase or urea (and other input prices) must decrease or both Which should it be

Table 3 provides some insight into this policy questionCurrent prices have been converted into real prices through the consumer price index (CPI) In real terms wheat and basmati prices have decreased over 20 since 1979 while IIRI-6 has only gone down in the past two years Real urea prices went upsubstantially in the early 1980s but are now down to roughly the 1979 level

The lower part of Table 3 carries the analysis further bydividing per capita GDP by the real prices Thus for example per capita GDP would purchase 531 kg of wheat in 1979 but 751 kgin 1985 a 42 increase as shown in the lower right-hand columns Only urea is at 1979 parity by this index While the ability to purchase much more food represents a substantial gainin consumers surplus over this period it seems that the processhas gone too far and some of the surplus should be reallocated back to producers as in the increased price support for wheat this year

CONCLUSIONS

It seems clear that fertilizer use and HYVs drive wheat production and that fertilizer use is in turn mainly driven bythe combination of expanded irrigated area and getting prices

6

Figure 3

NH E-T

21 ~

1 -I

-V7-4-4

i9-71 19-73 1975 1917 1BI 1B3 1

2- 4 SUP RET T-1 S SUP COST L

__- G _

9-7

Figure 4

AEAT

1

-4

- a

K

1971

WHLI-7 RET -

197- 3

E

4 -

C4-7 shy

HSL W

9-

ZC OT Tshy

191

-

1983 1985

UREAirHA

---------------------------------------------------------------------------------

--------------------------------------------------------------

TABLE 3PRICES AND INCOME

YEAR SUPPORTMARKET PRICES PER K6 REAL PRICES 1979 BASE WHEAT BASMATI IRRI-6 UREA CPI BASE WHEAT BASMATI IRRI-6 UREA

1979 121 161 081 128 296 1 121 161 081 128 1980 125 161 081 186 332 112 111 144 072 166 1981 145 188 097 192 375 127 114 148 077 152 1982 145 213 113 236 404 136 106 156 0P3 173 1983 16 22 123 256 418 141 113 156 087 181 19B4 16 225 128 256 462 156 103 144 082 164 1985 175 225 128 256 508 172 102 131 075 144 1986 2 233 133 256 559 1I9 106 123 070 136

YEAR RE4L PRICES INRELATION TO PE CAPITA GDP PERENT CHANGE OVE Fq73 (RUPEES 1959-1960 BASE EAL 6DFI PRICE PC GDF WHEAT BASMATI Ii-o UREA PC GDP WHEAT BASMATI IRRI-6 UREA

1975 642 531 39 73 502 100 10 100 100 100 19CC 666 598 464 922 402 104 113 116 116 0BF 1981 682 596 460 891 45) 106 112 115 112 090 1982 737 694 472 890 426 115 131 118 112 085 1983 736 650 472 845 406 115 122 118 107 081 1984 736 718 511 89- 449 115 135 128 113 089 1985 766 751 584 1027 514 119 142 147 130 102

SOURCE WORLD BANK 1985

right Subject to more detailed research than has been possiblehere several major policy issues emerge from this analysis

1) The development of new irrigated area seemed to have stopped in favor of rehabilitation of existing systems But simple physical rehabilitation only preserves the status quoante--in fact if not in theory--and thus the prime driver of Pakistans wheat production expansion of irrigated area has been stopped in its tracks

2) It is often thought that vertical production--iehigher yields--is the most economical path to agriculturaldevelopment This is a theory that has yet to be proven one about which the writer has serious reservations But even sovertical production does not come out of the air it comes out of inputs--of fertilizer HYVs and labor And the farmer must have the proper prices to encourage him to invest in these inputs It appears at least up to this year that agricultural policy in Pakistan has not been providing the proper price incentives and that crop prices should be increased Even substantial increases will leave consumers much better off than they were in 1979 relative to their income

Thus it appears to the writer that current fertilizer pricesshould be frozen for a year or two and that crop pricesespecially of wheat and basmati should be increased This mayprovide a short-run stimulus to crop production in Pakistan But in the long run the prospects for sustained growth without expansion of irrigated area and improved management of existingirrigation systems are dim Pakistan clearly needs to develop anew irrigation strategy based on an objective assessment of the experience of the past and careful quantitative assessments of the technical and managerial opportunities of the future

Finally unless Pakistan can control the growth of its population there is no foreseeable solution to the food problem

The author is Professor Department of Agricultural and Resource Economics Executive Director of the International School for Agricultural and Resource Development (ISARD) and Co-Director of the Colorado Institute of Irrigation Management (CIIM) at Colorado State University

This is Part A of a report to USAID Pakistan under a contract with Chemonics International Consulting Division

7

Appendix A WHEAT AND RICE RESPONSE TO RAINFALL IN THE INDUS BASIN (A TIME SERIES ANALYSIS) by Dr Zakir Hussain (USAID)

Wheat and rice are the two major food grain crops Of the total wheat

80 percent is cultivated in the irrigated areas and rice is almost

irrigated crop Both these crops experienced Green Revolution

technologies in the past two decades In the past two years

production of these crops is declining to an alarming extent Several

experts are attributirng this phenomena to unfavourable weather

condition In this note efforts have been made to help scan the

impact of rainfall in the critical months of growing season of rice

and wheat The ensuing discussion seperatly deals with the wheat and

rice

WHEAT PRODUCTIION RESPONSE TO RAINFALL

Wheat production and rainfall data was obtained from the Ministry of

food and agriculture for the past fifteen years (1971-1985)The

critical rainfall months December through January for wheat crop were

identified The same months were then deflated with the normal

rainfall of the these months ( normal rainfall is the average of

preceding twenty years) and eventualiy rainfall indicies were

calculated In similar fashion these indicies were identified as

independent variables for the provinces and overall equations The

equations are as follows

Y = a + bXi + 1 + e (1)

LnY= - + b LnXi + T + e (2) (i---ll)

Where

Y = production (000 tons) X = weather index Ln= Loge =

error termT= time trenda and b are parameters to be estimated

1

On Pakistan basis seven independent variables ( PunjabSind and

Pakistan linear and log linear rain indicies and trend) were

regressed against linear and log linear production of wheat Stepwis

regression was used to delineate the effect of each regressorThe

results are obtained in table 1 The analysis revealed that wheat

production has grown at 4 percent per annum and the coefficient of

trend variable was highly significant at 99 percent probability

level Of all the indicies Sind rainfall index turned out to be

significant perhaps it precipitates in the month of December in Sind

which is critical period for wheat crop tillering This will become

more clear from the foregoing analysisThere seems to be no problem

of auto-correlation (DW23) and explained variance is as high as 9

percent (R2 9)

The above two equations were fitted to Funjab data in log and

linear terms and stepwise regression was applied to help scan the

impact of rainfall in the critical months The results are again

reported in table 1 which shown that production grew at 3 percent

and January rain turned out to be critical month for this staple foot

crop Surprisingly enough the canal authorities close the canals in

the month of January which adversely effects the yield of wheat

There is no problem of auto-correlation (DW256) and R2 is fairly

high (93) In case of Sind provinc-e trend variable turned to be

highly significant and wheat production grows at annual rate of 5

percent Fuerthermore December rain has important bearing on the

productivity of this cropIt may nct bE out of place to mention that

wheat is sown 2-3 weeks ealier in Sind an compared to Funjab The

general fit of the equation is pretty good

--------------------------------------------------------------------------------------------------

--------------------------------------------------------------------------------------------------

TABLE 2 WHEAT AND UREA PRICES

YEAR YEAR A B r D E F G H I J SUPPORT WHOLESALE UREA WHEAT SUPPORT SUPPORT WHOLESALE WHOLESALE UREA USE WHEAT PRICE PRICE PRICE PROD PER WHEAT RSRETURN WHEAT RSRETURN INWHEAT YEILD WHEAT WHEAT KG UREA COST UREA WHEATUREA COST UREA WHEATUREA RSIKG RSIKGi RSiKG KG46m (CA) (DIE) (CB) (DI) KGHA KGHA

---------------------------------------------------------------------------shy

1971 1771 046 C57 35 124 287 244 108 172 1972 37 124 306 21 11991)44 5 1 197 973 6 117C7 34 280 299 14 1974 1974 062 114 304 16A 161 328 1248 1975 1975 09 117 11 419 111 377 11B 355 51 1320 1976 1976 099 104 148 245 149 164 105 234 465 1422 1977 1977 099 111 136 254 137 185 112 226 451 1431 1978 1578 099 137 136 212 137 154 138 153 497 1316 1979 1979 121 135 128 192 106 182 111 173 620 1488 198) 1960 125 136 186 168 149 113 108 155 745 156P 1961 1981 145 155 192 179 132 15 107 168 733 1647 1982 1992 145 18 26 18c 163 n 12 145 1565 192 19t 16 193 25 174 166 10 12 145 77) 167i 144 IPE 16 21 25 148 160 092 125 118 803 142219p) 1985 175 256 156 146 107 827 1612 1986 19B6 2 256 167 128 130 843 1757

SUPOT PRICES FOP THE CROF YEAR IE 1986 1985-1986) PRICES INLAHORE CALANDEF YEAR

The consequence of decreasing returns to investment in urea is illustrated in Figure 3 Here urea application per ha (divided by 10) is related to the (support price) return variable (F of Table 2) and to the wheat price of urea in the preceding year This figure indicates that as long as the wheat return is about 16 kg or more farmers will invest in fertilizer but below that amount they are reluctant to do so With the low returns of the past five years the question is not why the rate of growth of fertilizer use--and consequently of wheat production--slowed but rather why it grew at all

Figure 4 indicates that a possible answer to this question may lie in wholesale rather than support wheat prices The wheat cost of urea in terms of wholesale prices has been more stable than the support price index Many regression analyses were made to incorporate these price effects into the fertilizer functions of Table 1 but the results were not wholly satisfactory Much more detailed analysis is needed on these relationships than was possible here A promising line of further research on this subject is shown in R4 of Table 1 Here the change in urea use per ha is related to the change in the (support price) return to urea in the preceding year An R2 value of 051 is quite good for change variables

Nevertheless it seems clear that Pakistani farmers are going to need more favorable price policies if they are going to invest in increased production Either crop prices must increase or urea (and other input prices) must decrease or both Which should it be

Table 3 provides some insight into this policy questionCurrent prices have been converted into real prices through the consumer price index (CPI) In real terms wheat and basmati prices have decreased over 20 since 1979 while IIRI-6 has only gone down in the past two years Real urea prices went upsubstantially in the early 1980s but are now down to roughly the 1979 level

The lower part of Table 3 carries the analysis further bydividing per capita GDP by the real prices Thus for example per capita GDP would purchase 531 kg of wheat in 1979 but 751 kgin 1985 a 42 increase as shown in the lower right-hand columns Only urea is at 1979 parity by this index While the ability to purchase much more food represents a substantial gainin consumers surplus over this period it seems that the processhas gone too far and some of the surplus should be reallocated back to producers as in the increased price support for wheat this year

CONCLUSIONS

It seems clear that fertilizer use and HYVs drive wheat production and that fertilizer use is in turn mainly driven bythe combination of expanded irrigated area and getting prices

6

Figure 3

NH E-T

21 ~

1 -I

-V7-4-4

i9-71 19-73 1975 1917 1BI 1B3 1

2- 4 SUP RET T-1 S SUP COST L

__- G _

9-7

Figure 4

AEAT

1

-4

- a

K

1971

WHLI-7 RET -

197- 3

E

4 -

C4-7 shy

HSL W

9-

ZC OT Tshy

191

-

1983 1985

UREAirHA

---------------------------------------------------------------------------------

--------------------------------------------------------------

TABLE 3PRICES AND INCOME

YEAR SUPPORTMARKET PRICES PER K6 REAL PRICES 1979 BASE WHEAT BASMATI IRRI-6 UREA CPI BASE WHEAT BASMATI IRRI-6 UREA

1979 121 161 081 128 296 1 121 161 081 128 1980 125 161 081 186 332 112 111 144 072 166 1981 145 188 097 192 375 127 114 148 077 152 1982 145 213 113 236 404 136 106 156 0P3 173 1983 16 22 123 256 418 141 113 156 087 181 19B4 16 225 128 256 462 156 103 144 082 164 1985 175 225 128 256 508 172 102 131 075 144 1986 2 233 133 256 559 1I9 106 123 070 136

YEAR RE4L PRICES INRELATION TO PE CAPITA GDP PERENT CHANGE OVE Fq73 (RUPEES 1959-1960 BASE EAL 6DFI PRICE PC GDF WHEAT BASMATI Ii-o UREA PC GDP WHEAT BASMATI IRRI-6 UREA

1975 642 531 39 73 502 100 10 100 100 100 19CC 666 598 464 922 402 104 113 116 116 0BF 1981 682 596 460 891 45) 106 112 115 112 090 1982 737 694 472 890 426 115 131 118 112 085 1983 736 650 472 845 406 115 122 118 107 081 1984 736 718 511 89- 449 115 135 128 113 089 1985 766 751 584 1027 514 119 142 147 130 102

SOURCE WORLD BANK 1985

right Subject to more detailed research than has been possiblehere several major policy issues emerge from this analysis

1) The development of new irrigated area seemed to have stopped in favor of rehabilitation of existing systems But simple physical rehabilitation only preserves the status quoante--in fact if not in theory--and thus the prime driver of Pakistans wheat production expansion of irrigated area has been stopped in its tracks

2) It is often thought that vertical production--iehigher yields--is the most economical path to agriculturaldevelopment This is a theory that has yet to be proven one about which the writer has serious reservations But even sovertical production does not come out of the air it comes out of inputs--of fertilizer HYVs and labor And the farmer must have the proper prices to encourage him to invest in these inputs It appears at least up to this year that agricultural policy in Pakistan has not been providing the proper price incentives and that crop prices should be increased Even substantial increases will leave consumers much better off than they were in 1979 relative to their income

Thus it appears to the writer that current fertilizer pricesshould be frozen for a year or two and that crop pricesespecially of wheat and basmati should be increased This mayprovide a short-run stimulus to crop production in Pakistan But in the long run the prospects for sustained growth without expansion of irrigated area and improved management of existingirrigation systems are dim Pakistan clearly needs to develop anew irrigation strategy based on an objective assessment of the experience of the past and careful quantitative assessments of the technical and managerial opportunities of the future

Finally unless Pakistan can control the growth of its population there is no foreseeable solution to the food problem

The author is Professor Department of Agricultural and Resource Economics Executive Director of the International School for Agricultural and Resource Development (ISARD) and Co-Director of the Colorado Institute of Irrigation Management (CIIM) at Colorado State University

This is Part A of a report to USAID Pakistan under a contract with Chemonics International Consulting Division

7

Appendix A WHEAT AND RICE RESPONSE TO RAINFALL IN THE INDUS BASIN (A TIME SERIES ANALYSIS) by Dr Zakir Hussain (USAID)

Wheat and rice are the two major food grain crops Of the total wheat

80 percent is cultivated in the irrigated areas and rice is almost

irrigated crop Both these crops experienced Green Revolution

technologies in the past two decades In the past two years

production of these crops is declining to an alarming extent Several

experts are attributirng this phenomena to unfavourable weather

condition In this note efforts have been made to help scan the

impact of rainfall in the critical months of growing season of rice

and wheat The ensuing discussion seperatly deals with the wheat and

rice

WHEAT PRODUCTIION RESPONSE TO RAINFALL

Wheat production and rainfall data was obtained from the Ministry of

food and agriculture for the past fifteen years (1971-1985)The

critical rainfall months December through January for wheat crop were

identified The same months were then deflated with the normal

rainfall of the these months ( normal rainfall is the average of

preceding twenty years) and eventualiy rainfall indicies were

calculated In similar fashion these indicies were identified as

independent variables for the provinces and overall equations The

equations are as follows

Y = a + bXi + 1 + e (1)

LnY= - + b LnXi + T + e (2) (i---ll)

Where

Y = production (000 tons) X = weather index Ln= Loge =

error termT= time trenda and b are parameters to be estimated

1

On Pakistan basis seven independent variables ( PunjabSind and

Pakistan linear and log linear rain indicies and trend) were

regressed against linear and log linear production of wheat Stepwis

regression was used to delineate the effect of each regressorThe

results are obtained in table 1 The analysis revealed that wheat

production has grown at 4 percent per annum and the coefficient of

trend variable was highly significant at 99 percent probability

level Of all the indicies Sind rainfall index turned out to be

significant perhaps it precipitates in the month of December in Sind

which is critical period for wheat crop tillering This will become

more clear from the foregoing analysisThere seems to be no problem

of auto-correlation (DW23) and explained variance is as high as 9

percent (R2 9)

The above two equations were fitted to Funjab data in log and

linear terms and stepwise regression was applied to help scan the

impact of rainfall in the critical months The results are again

reported in table 1 which shown that production grew at 3 percent

and January rain turned out to be critical month for this staple foot

crop Surprisingly enough the canal authorities close the canals in

the month of January which adversely effects the yield of wheat

There is no problem of auto-correlation (DW256) and R2 is fairly

high (93) In case of Sind provinc-e trend variable turned to be

highly significant and wheat production grows at annual rate of 5

percent Fuerthermore December rain has important bearing on the

productivity of this cropIt may nct bE out of place to mention that

wheat is sown 2-3 weeks ealier in Sind an compared to Funjab The

general fit of the equation is pretty good

The consequence of decreasing returns to investment in urea is illustrated in Figure 3 Here urea application per ha (divided by 10) is related to the (support price) return variable (F of Table 2) and to the wheat price of urea in the preceding year This figure indicates that as long as the wheat return is about 16 kg or more farmers will invest in fertilizer but below that amount they are reluctant to do so With the low returns of the past five years the question is not why the rate of growth of fertilizer use--and consequently of wheat production--slowed but rather why it grew at all

Figure 4 indicates that a possible answer to this question may lie in wholesale rather than support wheat prices The wheat cost of urea in terms of wholesale prices has been more stable than the support price index Many regression analyses were made to incorporate these price effects into the fertilizer functions of Table 1 but the results were not wholly satisfactory Much more detailed analysis is needed on these relationships than was possible here A promising line of further research on this subject is shown in R4 of Table 1 Here the change in urea use per ha is related to the change in the (support price) return to urea in the preceding year An R2 value of 051 is quite good for change variables

Nevertheless it seems clear that Pakistani farmers are going to need more favorable price policies if they are going to invest in increased production Either crop prices must increase or urea (and other input prices) must decrease or both Which should it be

Table 3 provides some insight into this policy questionCurrent prices have been converted into real prices through the consumer price index (CPI) In real terms wheat and basmati prices have decreased over 20 since 1979 while IIRI-6 has only gone down in the past two years Real urea prices went upsubstantially in the early 1980s but are now down to roughly the 1979 level

The lower part of Table 3 carries the analysis further bydividing per capita GDP by the real prices Thus for example per capita GDP would purchase 531 kg of wheat in 1979 but 751 kgin 1985 a 42 increase as shown in the lower right-hand columns Only urea is at 1979 parity by this index While the ability to purchase much more food represents a substantial gainin consumers surplus over this period it seems that the processhas gone too far and some of the surplus should be reallocated back to producers as in the increased price support for wheat this year

CONCLUSIONS

It seems clear that fertilizer use and HYVs drive wheat production and that fertilizer use is in turn mainly driven bythe combination of expanded irrigated area and getting prices

6

Figure 3

NH E-T

21 ~

1 -I

-V7-4-4

i9-71 19-73 1975 1917 1BI 1B3 1

2- 4 SUP RET T-1 S SUP COST L

__- G _

9-7

Figure 4

AEAT

1

-4

- a

K

1971

WHLI-7 RET -

197- 3

E

4 -

C4-7 shy

HSL W

9-

ZC OT Tshy

191

-

1983 1985

UREAirHA

---------------------------------------------------------------------------------

--------------------------------------------------------------

TABLE 3PRICES AND INCOME

YEAR SUPPORTMARKET PRICES PER K6 REAL PRICES 1979 BASE WHEAT BASMATI IRRI-6 UREA CPI BASE WHEAT BASMATI IRRI-6 UREA

1979 121 161 081 128 296 1 121 161 081 128 1980 125 161 081 186 332 112 111 144 072 166 1981 145 188 097 192 375 127 114 148 077 152 1982 145 213 113 236 404 136 106 156 0P3 173 1983 16 22 123 256 418 141 113 156 087 181 19B4 16 225 128 256 462 156 103 144 082 164 1985 175 225 128 256 508 172 102 131 075 144 1986 2 233 133 256 559 1I9 106 123 070 136

YEAR RE4L PRICES INRELATION TO PE CAPITA GDP PERENT CHANGE OVE Fq73 (RUPEES 1959-1960 BASE EAL 6DFI PRICE PC GDF WHEAT BASMATI Ii-o UREA PC GDP WHEAT BASMATI IRRI-6 UREA

1975 642 531 39 73 502 100 10 100 100 100 19CC 666 598 464 922 402 104 113 116 116 0BF 1981 682 596 460 891 45) 106 112 115 112 090 1982 737 694 472 890 426 115 131 118 112 085 1983 736 650 472 845 406 115 122 118 107 081 1984 736 718 511 89- 449 115 135 128 113 089 1985 766 751 584 1027 514 119 142 147 130 102

SOURCE WORLD BANK 1985

right Subject to more detailed research than has been possiblehere several major policy issues emerge from this analysis

1) The development of new irrigated area seemed to have stopped in favor of rehabilitation of existing systems But simple physical rehabilitation only preserves the status quoante--in fact if not in theory--and thus the prime driver of Pakistans wheat production expansion of irrigated area has been stopped in its tracks

2) It is often thought that vertical production--iehigher yields--is the most economical path to agriculturaldevelopment This is a theory that has yet to be proven one about which the writer has serious reservations But even sovertical production does not come out of the air it comes out of inputs--of fertilizer HYVs and labor And the farmer must have the proper prices to encourage him to invest in these inputs It appears at least up to this year that agricultural policy in Pakistan has not been providing the proper price incentives and that crop prices should be increased Even substantial increases will leave consumers much better off than they were in 1979 relative to their income

Thus it appears to the writer that current fertilizer pricesshould be frozen for a year or two and that crop pricesespecially of wheat and basmati should be increased This mayprovide a short-run stimulus to crop production in Pakistan But in the long run the prospects for sustained growth without expansion of irrigated area and improved management of existingirrigation systems are dim Pakistan clearly needs to develop anew irrigation strategy based on an objective assessment of the experience of the past and careful quantitative assessments of the technical and managerial opportunities of the future

Finally unless Pakistan can control the growth of its population there is no foreseeable solution to the food problem

The author is Professor Department of Agricultural and Resource Economics Executive Director of the International School for Agricultural and Resource Development (ISARD) and Co-Director of the Colorado Institute of Irrigation Management (CIIM) at Colorado State University

This is Part A of a report to USAID Pakistan under a contract with Chemonics International Consulting Division

7

Appendix A WHEAT AND RICE RESPONSE TO RAINFALL IN THE INDUS BASIN (A TIME SERIES ANALYSIS) by Dr Zakir Hussain (USAID)

Wheat and rice are the two major food grain crops Of the total wheat

80 percent is cultivated in the irrigated areas and rice is almost

irrigated crop Both these crops experienced Green Revolution

technologies in the past two decades In the past two years

production of these crops is declining to an alarming extent Several

experts are attributirng this phenomena to unfavourable weather

condition In this note efforts have been made to help scan the

impact of rainfall in the critical months of growing season of rice

and wheat The ensuing discussion seperatly deals with the wheat and

rice

WHEAT PRODUCTIION RESPONSE TO RAINFALL

Wheat production and rainfall data was obtained from the Ministry of

food and agriculture for the past fifteen years (1971-1985)The

critical rainfall months December through January for wheat crop were

identified The same months were then deflated with the normal

rainfall of the these months ( normal rainfall is the average of

preceding twenty years) and eventualiy rainfall indicies were

calculated In similar fashion these indicies were identified as

independent variables for the provinces and overall equations The

equations are as follows

Y = a + bXi + 1 + e (1)

LnY= - + b LnXi + T + e (2) (i---ll)

Where

Y = production (000 tons) X = weather index Ln= Loge =

error termT= time trenda and b are parameters to be estimated

1

On Pakistan basis seven independent variables ( PunjabSind and

Pakistan linear and log linear rain indicies and trend) were

regressed against linear and log linear production of wheat Stepwis

regression was used to delineate the effect of each regressorThe

results are obtained in table 1 The analysis revealed that wheat

production has grown at 4 percent per annum and the coefficient of

trend variable was highly significant at 99 percent probability

level Of all the indicies Sind rainfall index turned out to be

significant perhaps it precipitates in the month of December in Sind

which is critical period for wheat crop tillering This will become

more clear from the foregoing analysisThere seems to be no problem

of auto-correlation (DW23) and explained variance is as high as 9

percent (R2 9)

The above two equations were fitted to Funjab data in log and

linear terms and stepwise regression was applied to help scan the

impact of rainfall in the critical months The results are again

reported in table 1 which shown that production grew at 3 percent

and January rain turned out to be critical month for this staple foot

crop Surprisingly enough the canal authorities close the canals in

the month of January which adversely effects the yield of wheat

There is no problem of auto-correlation (DW256) and R2 is fairly

high (93) In case of Sind provinc-e trend variable turned to be

highly significant and wheat production grows at annual rate of 5

percent Fuerthermore December rain has important bearing on the

productivity of this cropIt may nct bE out of place to mention that

wheat is sown 2-3 weeks ealier in Sind an compared to Funjab The

general fit of the equation is pretty good

Figure 3

NH E-T

21 ~

1 -I

-V7-4-4

i9-71 19-73 1975 1917 1BI 1B3 1

2- 4 SUP RET T-1 S SUP COST L

__- G _

9-7

Figure 4

AEAT

1

-4

- a

K

1971

WHLI-7 RET -

197- 3

E

4 -

C4-7 shy

HSL W

9-

ZC OT Tshy

191

-

1983 1985

UREAirHA

---------------------------------------------------------------------------------

--------------------------------------------------------------

TABLE 3PRICES AND INCOME

YEAR SUPPORTMARKET PRICES PER K6 REAL PRICES 1979 BASE WHEAT BASMATI IRRI-6 UREA CPI BASE WHEAT BASMATI IRRI-6 UREA

1979 121 161 081 128 296 1 121 161 081 128 1980 125 161 081 186 332 112 111 144 072 166 1981 145 188 097 192 375 127 114 148 077 152 1982 145 213 113 236 404 136 106 156 0P3 173 1983 16 22 123 256 418 141 113 156 087 181 19B4 16 225 128 256 462 156 103 144 082 164 1985 175 225 128 256 508 172 102 131 075 144 1986 2 233 133 256 559 1I9 106 123 070 136

YEAR RE4L PRICES INRELATION TO PE CAPITA GDP PERENT CHANGE OVE Fq73 (RUPEES 1959-1960 BASE EAL 6DFI PRICE PC GDF WHEAT BASMATI Ii-o UREA PC GDP WHEAT BASMATI IRRI-6 UREA

1975 642 531 39 73 502 100 10 100 100 100 19CC 666 598 464 922 402 104 113 116 116 0BF 1981 682 596 460 891 45) 106 112 115 112 090 1982 737 694 472 890 426 115 131 118 112 085 1983 736 650 472 845 406 115 122 118 107 081 1984 736 718 511 89- 449 115 135 128 113 089 1985 766 751 584 1027 514 119 142 147 130 102

SOURCE WORLD BANK 1985

right Subject to more detailed research than has been possiblehere several major policy issues emerge from this analysis

1) The development of new irrigated area seemed to have stopped in favor of rehabilitation of existing systems But simple physical rehabilitation only preserves the status quoante--in fact if not in theory--and thus the prime driver of Pakistans wheat production expansion of irrigated area has been stopped in its tracks

2) It is often thought that vertical production--iehigher yields--is the most economical path to agriculturaldevelopment This is a theory that has yet to be proven one about which the writer has serious reservations But even sovertical production does not come out of the air it comes out of inputs--of fertilizer HYVs and labor And the farmer must have the proper prices to encourage him to invest in these inputs It appears at least up to this year that agricultural policy in Pakistan has not been providing the proper price incentives and that crop prices should be increased Even substantial increases will leave consumers much better off than they were in 1979 relative to their income

Thus it appears to the writer that current fertilizer pricesshould be frozen for a year or two and that crop pricesespecially of wheat and basmati should be increased This mayprovide a short-run stimulus to crop production in Pakistan But in the long run the prospects for sustained growth without expansion of irrigated area and improved management of existingirrigation systems are dim Pakistan clearly needs to develop anew irrigation strategy based on an objective assessment of the experience of the past and careful quantitative assessments of the technical and managerial opportunities of the future

Finally unless Pakistan can control the growth of its population there is no foreseeable solution to the food problem

The author is Professor Department of Agricultural and Resource Economics Executive Director of the International School for Agricultural and Resource Development (ISARD) and Co-Director of the Colorado Institute of Irrigation Management (CIIM) at Colorado State University

This is Part A of a report to USAID Pakistan under a contract with Chemonics International Consulting Division

7

Appendix A WHEAT AND RICE RESPONSE TO RAINFALL IN THE INDUS BASIN (A TIME SERIES ANALYSIS) by Dr Zakir Hussain (USAID)

Wheat and rice are the two major food grain crops Of the total wheat

80 percent is cultivated in the irrigated areas and rice is almost

irrigated crop Both these crops experienced Green Revolution

technologies in the past two decades In the past two years

production of these crops is declining to an alarming extent Several

experts are attributirng this phenomena to unfavourable weather

condition In this note efforts have been made to help scan the

impact of rainfall in the critical months of growing season of rice

and wheat The ensuing discussion seperatly deals with the wheat and

rice

WHEAT PRODUCTIION RESPONSE TO RAINFALL

Wheat production and rainfall data was obtained from the Ministry of

food and agriculture for the past fifteen years (1971-1985)The

critical rainfall months December through January for wheat crop were

identified The same months were then deflated with the normal

rainfall of the these months ( normal rainfall is the average of

preceding twenty years) and eventualiy rainfall indicies were

calculated In similar fashion these indicies were identified as

independent variables for the provinces and overall equations The

equations are as follows

Y = a + bXi + 1 + e (1)

LnY= - + b LnXi + T + e (2) (i---ll)

Where

Y = production (000 tons) X = weather index Ln= Loge =

error termT= time trenda and b are parameters to be estimated

1

On Pakistan basis seven independent variables ( PunjabSind and

Pakistan linear and log linear rain indicies and trend) were

regressed against linear and log linear production of wheat Stepwis

regression was used to delineate the effect of each regressorThe

results are obtained in table 1 The analysis revealed that wheat

production has grown at 4 percent per annum and the coefficient of

trend variable was highly significant at 99 percent probability

level Of all the indicies Sind rainfall index turned out to be

significant perhaps it precipitates in the month of December in Sind

which is critical period for wheat crop tillering This will become

more clear from the foregoing analysisThere seems to be no problem

of auto-correlation (DW23) and explained variance is as high as 9

percent (R2 9)

The above two equations were fitted to Funjab data in log and

linear terms and stepwise regression was applied to help scan the

impact of rainfall in the critical months The results are again

reported in table 1 which shown that production grew at 3 percent

and January rain turned out to be critical month for this staple foot

crop Surprisingly enough the canal authorities close the canals in

the month of January which adversely effects the yield of wheat

There is no problem of auto-correlation (DW256) and R2 is fairly

high (93) In case of Sind provinc-e trend variable turned to be

highly significant and wheat production grows at annual rate of 5

percent Fuerthermore December rain has important bearing on the

productivity of this cropIt may nct bE out of place to mention that

wheat is sown 2-3 weeks ealier in Sind an compared to Funjab The

general fit of the equation is pretty good

9-7

Figure 4

AEAT

1

-4

- a

K

1971

WHLI-7 RET -

197- 3

E

4 -

C4-7 shy

HSL W

9-

ZC OT Tshy

191

-

1983 1985

UREAirHA

---------------------------------------------------------------------------------

--------------------------------------------------------------

TABLE 3PRICES AND INCOME

YEAR SUPPORTMARKET PRICES PER K6 REAL PRICES 1979 BASE WHEAT BASMATI IRRI-6 UREA CPI BASE WHEAT BASMATI IRRI-6 UREA

1979 121 161 081 128 296 1 121 161 081 128 1980 125 161 081 186 332 112 111 144 072 166 1981 145 188 097 192 375 127 114 148 077 152 1982 145 213 113 236 404 136 106 156 0P3 173 1983 16 22 123 256 418 141 113 156 087 181 19B4 16 225 128 256 462 156 103 144 082 164 1985 175 225 128 256 508 172 102 131 075 144 1986 2 233 133 256 559 1I9 106 123 070 136

YEAR RE4L PRICES INRELATION TO PE CAPITA GDP PERENT CHANGE OVE Fq73 (RUPEES 1959-1960 BASE EAL 6DFI PRICE PC GDF WHEAT BASMATI Ii-o UREA PC GDP WHEAT BASMATI IRRI-6 UREA

1975 642 531 39 73 502 100 10 100 100 100 19CC 666 598 464 922 402 104 113 116 116 0BF 1981 682 596 460 891 45) 106 112 115 112 090 1982 737 694 472 890 426 115 131 118 112 085 1983 736 650 472 845 406 115 122 118 107 081 1984 736 718 511 89- 449 115 135 128 113 089 1985 766 751 584 1027 514 119 142 147 130 102

SOURCE WORLD BANK 1985

right Subject to more detailed research than has been possiblehere several major policy issues emerge from this analysis

1) The development of new irrigated area seemed to have stopped in favor of rehabilitation of existing systems But simple physical rehabilitation only preserves the status quoante--in fact if not in theory--and thus the prime driver of Pakistans wheat production expansion of irrigated area has been stopped in its tracks

2) It is often thought that vertical production--iehigher yields--is the most economical path to agriculturaldevelopment This is a theory that has yet to be proven one about which the writer has serious reservations But even sovertical production does not come out of the air it comes out of inputs--of fertilizer HYVs and labor And the farmer must have the proper prices to encourage him to invest in these inputs It appears at least up to this year that agricultural policy in Pakistan has not been providing the proper price incentives and that crop prices should be increased Even substantial increases will leave consumers much better off than they were in 1979 relative to their income

Thus it appears to the writer that current fertilizer pricesshould be frozen for a year or two and that crop pricesespecially of wheat and basmati should be increased This mayprovide a short-run stimulus to crop production in Pakistan But in the long run the prospects for sustained growth without expansion of irrigated area and improved management of existingirrigation systems are dim Pakistan clearly needs to develop anew irrigation strategy based on an objective assessment of the experience of the past and careful quantitative assessments of the technical and managerial opportunities of the future

Finally unless Pakistan can control the growth of its population there is no foreseeable solution to the food problem

The author is Professor Department of Agricultural and Resource Economics Executive Director of the International School for Agricultural and Resource Development (ISARD) and Co-Director of the Colorado Institute of Irrigation Management (CIIM) at Colorado State University

This is Part A of a report to USAID Pakistan under a contract with Chemonics International Consulting Division

7

Appendix A WHEAT AND RICE RESPONSE TO RAINFALL IN THE INDUS BASIN (A TIME SERIES ANALYSIS) by Dr Zakir Hussain (USAID)

Wheat and rice are the two major food grain crops Of the total wheat

80 percent is cultivated in the irrigated areas and rice is almost

irrigated crop Both these crops experienced Green Revolution

technologies in the past two decades In the past two years

production of these crops is declining to an alarming extent Several

experts are attributirng this phenomena to unfavourable weather

condition In this note efforts have been made to help scan the

impact of rainfall in the critical months of growing season of rice

and wheat The ensuing discussion seperatly deals with the wheat and

rice

WHEAT PRODUCTIION RESPONSE TO RAINFALL

Wheat production and rainfall data was obtained from the Ministry of

food and agriculture for the past fifteen years (1971-1985)The

critical rainfall months December through January for wheat crop were

identified The same months were then deflated with the normal

rainfall of the these months ( normal rainfall is the average of

preceding twenty years) and eventualiy rainfall indicies were

calculated In similar fashion these indicies were identified as

independent variables for the provinces and overall equations The

equations are as follows

Y = a + bXi + 1 + e (1)

LnY= - + b LnXi + T + e (2) (i---ll)

Where

Y = production (000 tons) X = weather index Ln= Loge =

error termT= time trenda and b are parameters to be estimated

1

On Pakistan basis seven independent variables ( PunjabSind and

Pakistan linear and log linear rain indicies and trend) were

regressed against linear and log linear production of wheat Stepwis

regression was used to delineate the effect of each regressorThe

results are obtained in table 1 The analysis revealed that wheat

production has grown at 4 percent per annum and the coefficient of

trend variable was highly significant at 99 percent probability

level Of all the indicies Sind rainfall index turned out to be

significant perhaps it precipitates in the month of December in Sind

which is critical period for wheat crop tillering This will become

more clear from the foregoing analysisThere seems to be no problem

of auto-correlation (DW23) and explained variance is as high as 9

percent (R2 9)

The above two equations were fitted to Funjab data in log and

linear terms and stepwise regression was applied to help scan the

impact of rainfall in the critical months The results are again

reported in table 1 which shown that production grew at 3 percent

and January rain turned out to be critical month for this staple foot

crop Surprisingly enough the canal authorities close the canals in

the month of January which adversely effects the yield of wheat

There is no problem of auto-correlation (DW256) and R2 is fairly

high (93) In case of Sind provinc-e trend variable turned to be

highly significant and wheat production grows at annual rate of 5

percent Fuerthermore December rain has important bearing on the

productivity of this cropIt may nct bE out of place to mention that

wheat is sown 2-3 weeks ealier in Sind an compared to Funjab The

general fit of the equation is pretty good

---------------------------------------------------------------------------------

--------------------------------------------------------------

TABLE 3PRICES AND INCOME

YEAR SUPPORTMARKET PRICES PER K6 REAL PRICES 1979 BASE WHEAT BASMATI IRRI-6 UREA CPI BASE WHEAT BASMATI IRRI-6 UREA

1979 121 161 081 128 296 1 121 161 081 128 1980 125 161 081 186 332 112 111 144 072 166 1981 145 188 097 192 375 127 114 148 077 152 1982 145 213 113 236 404 136 106 156 0P3 173 1983 16 22 123 256 418 141 113 156 087 181 19B4 16 225 128 256 462 156 103 144 082 164 1985 175 225 128 256 508 172 102 131 075 144 1986 2 233 133 256 559 1I9 106 123 070 136

YEAR RE4L PRICES INRELATION TO PE CAPITA GDP PERENT CHANGE OVE Fq73 (RUPEES 1959-1960 BASE EAL 6DFI PRICE PC GDF WHEAT BASMATI Ii-o UREA PC GDP WHEAT BASMATI IRRI-6 UREA

1975 642 531 39 73 502 100 10 100 100 100 19CC 666 598 464 922 402 104 113 116 116 0BF 1981 682 596 460 891 45) 106 112 115 112 090 1982 737 694 472 890 426 115 131 118 112 085 1983 736 650 472 845 406 115 122 118 107 081 1984 736 718 511 89- 449 115 135 128 113 089 1985 766 751 584 1027 514 119 142 147 130 102

SOURCE WORLD BANK 1985

right Subject to more detailed research than has been possiblehere several major policy issues emerge from this analysis

1) The development of new irrigated area seemed to have stopped in favor of rehabilitation of existing systems But simple physical rehabilitation only preserves the status quoante--in fact if not in theory--and thus the prime driver of Pakistans wheat production expansion of irrigated area has been stopped in its tracks

2) It is often thought that vertical production--iehigher yields--is the most economical path to agriculturaldevelopment This is a theory that has yet to be proven one about which the writer has serious reservations But even sovertical production does not come out of the air it comes out of inputs--of fertilizer HYVs and labor And the farmer must have the proper prices to encourage him to invest in these inputs It appears at least up to this year that agricultural policy in Pakistan has not been providing the proper price incentives and that crop prices should be increased Even substantial increases will leave consumers much better off than they were in 1979 relative to their income

Thus it appears to the writer that current fertilizer pricesshould be frozen for a year or two and that crop pricesespecially of wheat and basmati should be increased This mayprovide a short-run stimulus to crop production in Pakistan But in the long run the prospects for sustained growth without expansion of irrigated area and improved management of existingirrigation systems are dim Pakistan clearly needs to develop anew irrigation strategy based on an objective assessment of the experience of the past and careful quantitative assessments of the technical and managerial opportunities of the future

Finally unless Pakistan can control the growth of its population there is no foreseeable solution to the food problem

The author is Professor Department of Agricultural and Resource Economics Executive Director of the International School for Agricultural and Resource Development (ISARD) and Co-Director of the Colorado Institute of Irrigation Management (CIIM) at Colorado State University

This is Part A of a report to USAID Pakistan under a contract with Chemonics International Consulting Division

7

Appendix A WHEAT AND RICE RESPONSE TO RAINFALL IN THE INDUS BASIN (A TIME SERIES ANALYSIS) by Dr Zakir Hussain (USAID)

Wheat and rice are the two major food grain crops Of the total wheat

80 percent is cultivated in the irrigated areas and rice is almost

irrigated crop Both these crops experienced Green Revolution

technologies in the past two decades In the past two years

production of these crops is declining to an alarming extent Several

experts are attributirng this phenomena to unfavourable weather

condition In this note efforts have been made to help scan the

impact of rainfall in the critical months of growing season of rice

and wheat The ensuing discussion seperatly deals with the wheat and

rice

WHEAT PRODUCTIION RESPONSE TO RAINFALL

Wheat production and rainfall data was obtained from the Ministry of

food and agriculture for the past fifteen years (1971-1985)The

critical rainfall months December through January for wheat crop were

identified The same months were then deflated with the normal

rainfall of the these months ( normal rainfall is the average of

preceding twenty years) and eventualiy rainfall indicies were

calculated In similar fashion these indicies were identified as

independent variables for the provinces and overall equations The

equations are as follows

Y = a + bXi + 1 + e (1)

LnY= - + b LnXi + T + e (2) (i---ll)

Where

Y = production (000 tons) X = weather index Ln= Loge =

error termT= time trenda and b are parameters to be estimated

1

On Pakistan basis seven independent variables ( PunjabSind and

Pakistan linear and log linear rain indicies and trend) were

regressed against linear and log linear production of wheat Stepwis

regression was used to delineate the effect of each regressorThe

results are obtained in table 1 The analysis revealed that wheat

production has grown at 4 percent per annum and the coefficient of

trend variable was highly significant at 99 percent probability

level Of all the indicies Sind rainfall index turned out to be

significant perhaps it precipitates in the month of December in Sind

which is critical period for wheat crop tillering This will become

more clear from the foregoing analysisThere seems to be no problem

of auto-correlation (DW23) and explained variance is as high as 9

percent (R2 9)

The above two equations were fitted to Funjab data in log and

linear terms and stepwise regression was applied to help scan the

impact of rainfall in the critical months The results are again

reported in table 1 which shown that production grew at 3 percent

and January rain turned out to be critical month for this staple foot

crop Surprisingly enough the canal authorities close the canals in

the month of January which adversely effects the yield of wheat

There is no problem of auto-correlation (DW256) and R2 is fairly

high (93) In case of Sind provinc-e trend variable turned to be

highly significant and wheat production grows at annual rate of 5

percent Fuerthermore December rain has important bearing on the

productivity of this cropIt may nct bE out of place to mention that

wheat is sown 2-3 weeks ealier in Sind an compared to Funjab The

general fit of the equation is pretty good

right Subject to more detailed research than has been possiblehere several major policy issues emerge from this analysis

1) The development of new irrigated area seemed to have stopped in favor of rehabilitation of existing systems But simple physical rehabilitation only preserves the status quoante--in fact if not in theory--and thus the prime driver of Pakistans wheat production expansion of irrigated area has been stopped in its tracks

2) It is often thought that vertical production--iehigher yields--is the most economical path to agriculturaldevelopment This is a theory that has yet to be proven one about which the writer has serious reservations But even sovertical production does not come out of the air it comes out of inputs--of fertilizer HYVs and labor And the farmer must have the proper prices to encourage him to invest in these inputs It appears at least up to this year that agricultural policy in Pakistan has not been providing the proper price incentives and that crop prices should be increased Even substantial increases will leave consumers much better off than they were in 1979 relative to their income

Thus it appears to the writer that current fertilizer pricesshould be frozen for a year or two and that crop pricesespecially of wheat and basmati should be increased This mayprovide a short-run stimulus to crop production in Pakistan But in the long run the prospects for sustained growth without expansion of irrigated area and improved management of existingirrigation systems are dim Pakistan clearly needs to develop anew irrigation strategy based on an objective assessment of the experience of the past and careful quantitative assessments of the technical and managerial opportunities of the future

Finally unless Pakistan can control the growth of its population there is no foreseeable solution to the food problem

The author is Professor Department of Agricultural and Resource Economics Executive Director of the International School for Agricultural and Resource Development (ISARD) and Co-Director of the Colorado Institute of Irrigation Management (CIIM) at Colorado State University

This is Part A of a report to USAID Pakistan under a contract with Chemonics International Consulting Division

7

Appendix A WHEAT AND RICE RESPONSE TO RAINFALL IN THE INDUS BASIN (A TIME SERIES ANALYSIS) by Dr Zakir Hussain (USAID)

Wheat and rice are the two major food grain crops Of the total wheat

80 percent is cultivated in the irrigated areas and rice is almost

irrigated crop Both these crops experienced Green Revolution

technologies in the past two decades In the past two years

production of these crops is declining to an alarming extent Several

experts are attributirng this phenomena to unfavourable weather

condition In this note efforts have been made to help scan the

impact of rainfall in the critical months of growing season of rice

and wheat The ensuing discussion seperatly deals with the wheat and

rice

WHEAT PRODUCTIION RESPONSE TO RAINFALL

Wheat production and rainfall data was obtained from the Ministry of

food and agriculture for the past fifteen years (1971-1985)The

critical rainfall months December through January for wheat crop were

identified The same months were then deflated with the normal

rainfall of the these months ( normal rainfall is the average of

preceding twenty years) and eventualiy rainfall indicies were

calculated In similar fashion these indicies were identified as

independent variables for the provinces and overall equations The

equations are as follows

Y = a + bXi + 1 + e (1)

LnY= - + b LnXi + T + e (2) (i---ll)

Where

Y = production (000 tons) X = weather index Ln= Loge =

error termT= time trenda and b are parameters to be estimated

1

On Pakistan basis seven independent variables ( PunjabSind and

Pakistan linear and log linear rain indicies and trend) were

regressed against linear and log linear production of wheat Stepwis

regression was used to delineate the effect of each regressorThe

results are obtained in table 1 The analysis revealed that wheat

production has grown at 4 percent per annum and the coefficient of

trend variable was highly significant at 99 percent probability

level Of all the indicies Sind rainfall index turned out to be

significant perhaps it precipitates in the month of December in Sind

which is critical period for wheat crop tillering This will become

more clear from the foregoing analysisThere seems to be no problem

of auto-correlation (DW23) and explained variance is as high as 9

percent (R2 9)

The above two equations were fitted to Funjab data in log and

linear terms and stepwise regression was applied to help scan the

impact of rainfall in the critical months The results are again

reported in table 1 which shown that production grew at 3 percent

and January rain turned out to be critical month for this staple foot

crop Surprisingly enough the canal authorities close the canals in

the month of January which adversely effects the yield of wheat

There is no problem of auto-correlation (DW256) and R2 is fairly

high (93) In case of Sind provinc-e trend variable turned to be

highly significant and wheat production grows at annual rate of 5

percent Fuerthermore December rain has important bearing on the

productivity of this cropIt may nct bE out of place to mention that

wheat is sown 2-3 weeks ealier in Sind an compared to Funjab The

general fit of the equation is pretty good

Appendix A WHEAT AND RICE RESPONSE TO RAINFALL IN THE INDUS BASIN (A TIME SERIES ANALYSIS) by Dr Zakir Hussain (USAID)

Wheat and rice are the two major food grain crops Of the total wheat

80 percent is cultivated in the irrigated areas and rice is almost

irrigated crop Both these crops experienced Green Revolution

technologies in the past two decades In the past two years

production of these crops is declining to an alarming extent Several

experts are attributirng this phenomena to unfavourable weather

condition In this note efforts have been made to help scan the

impact of rainfall in the critical months of growing season of rice

and wheat The ensuing discussion seperatly deals with the wheat and

rice

WHEAT PRODUCTIION RESPONSE TO RAINFALL

Wheat production and rainfall data was obtained from the Ministry of

food and agriculture for the past fifteen years (1971-1985)The

critical rainfall months December through January for wheat crop were

identified The same months were then deflated with the normal

rainfall of the these months ( normal rainfall is the average of

preceding twenty years) and eventualiy rainfall indicies were

calculated In similar fashion these indicies were identified as

independent variables for the provinces and overall equations The

equations are as follows

Y = a + bXi + 1 + e (1)

LnY= - + b LnXi + T + e (2) (i---ll)

Where

Y = production (000 tons) X = weather index Ln= Loge =

error termT= time trenda and b are parameters to be estimated

1

On Pakistan basis seven independent variables ( PunjabSind and

Pakistan linear and log linear rain indicies and trend) were

regressed against linear and log linear production of wheat Stepwis

regression was used to delineate the effect of each regressorThe

results are obtained in table 1 The analysis revealed that wheat

production has grown at 4 percent per annum and the coefficient of

trend variable was highly significant at 99 percent probability

level Of all the indicies Sind rainfall index turned out to be

significant perhaps it precipitates in the month of December in Sind

which is critical period for wheat crop tillering This will become

more clear from the foregoing analysisThere seems to be no problem

of auto-correlation (DW23) and explained variance is as high as 9

percent (R2 9)

The above two equations were fitted to Funjab data in log and

linear terms and stepwise regression was applied to help scan the

impact of rainfall in the critical months The results are again

reported in table 1 which shown that production grew at 3 percent

and January rain turned out to be critical month for this staple foot

crop Surprisingly enough the canal authorities close the canals in

the month of January which adversely effects the yield of wheat

There is no problem of auto-correlation (DW256) and R2 is fairly

high (93) In case of Sind provinc-e trend variable turned to be

highly significant and wheat production grows at annual rate of 5

percent Fuerthermore December rain has important bearing on the

productivity of this cropIt may nct bE out of place to mention that

wheat is sown 2-3 weeks ealier in Sind an compared to Funjab The

general fit of the equation is pretty good

On Pakistan basis seven independent variables ( PunjabSind and

Pakistan linear and log linear rain indicies and trend) were

regressed against linear and log linear production of wheat Stepwis

regression was used to delineate the effect of each regressorThe

results are obtained in table 1 The analysis revealed that wheat

production has grown at 4 percent per annum and the coefficient of

trend variable was highly significant at 99 percent probability

level Of all the indicies Sind rainfall index turned out to be

significant perhaps it precipitates in the month of December in Sind

which is critical period for wheat crop tillering This will become

more clear from the foregoing analysisThere seems to be no problem

of auto-correlation (DW23) and explained variance is as high as 9

percent (R2 9)

The above two equations were fitted to Funjab data in log and

linear terms and stepwise regression was applied to help scan the

impact of rainfall in the critical months The results are again

reported in table 1 which shown that production grew at 3 percent

and January rain turned out to be critical month for this staple foot

crop Surprisingly enough the canal authorities close the canals in

the month of January which adversely effects the yield of wheat

There is no problem of auto-correlation (DW256) and R2 is fairly

high (93) In case of Sind provinc-e trend variable turned to be

highly significant and wheat production grows at annual rate of 5

percent Fuerthermore December rain has important bearing on the

productivity of this cropIt may nct bE out of place to mention that

wheat is sown 2-3 weeks ealier in Sind an compared to Funjab The

general fit of the equation is pretty good

RICE PRODUCTION RESFONSE TO RAINFALL

Rice production in Pakistan is mainly irrigated

Howeverprecipitation does effect its productivity Due to monson

rainfall irrigation water in the canals is plentiful and rice tracts

also get augmented water suppliesThe critical months identified wer

JulyAugust and September As explained above similar rain indicies

for these months and repective proviIces were developed The above

mentioned two equation wure subjected to stepwise regression

analysis The overall Pakistan results show that rice production

grows at annual rate of 24 percent and non of the rainfall indicies

appeared in both the equations Sind results are similar to Fakistan

and production growth rate is 15 percent However the Punjab

situation is quite different The production trend is 28 percent and

August rain index linear as well as log term appeared in the

equations August rain has tonic effect on the rice yield but too

much rain in this month becomes toxic to crop due to pest

infestation ( wet weather) and waterlogging

----------------------------------------------------------------------------------------------

----------------------------------------------------------------------------------------------

Table 1

REGRESSION RESULTS RELATING WHEAT AND RICE PRODUCTION TO RAINFALL

Dependent Constant Trend Sind Rain January Rain Auqust Rain DEC Rain R2 Variable DW ---------------------------------------------------------------------

WHEAT PRODUCTION (000) TONS

Pakistan -735397 37628 257 092

(33309) (1222) (231)

Pakistan -74223 0042 000029 0

(0003) (000013) (231)

Punjab -490921 25153 8377 093 (22472) (2651) (255)

Punjab -65462 0038 0001 093 (0003) (0003) (242)

Sind -99903 0054 000009 094 (0004) (00005) (18)

RICE PRODUCTION (000) TONS

Pakistan -118904 61482 063 (12722) (057)

Pakistan -3B574 0023 064 (0004) (057)

Punjab -14297 35402 -14297 071 (7255) (62427) (136)

Punjab -49503 0028 -0662 0005 089 (0003) (0138) (0001) (131)

Sind -22552 0015 031 (0006) (079)

fI The coefficient issignificantly different from zero at 01 probability level

41 Log function

Standard error

1

Appendix B 11w Agricultra ELtzN of Pi lefal in a R gkAul Ir6ptrtlv

0I)m-19142)

K1ak- Sri1lhJ L 101h II1al LiduJh L~aamp

I Lruxi Aiva (000 2) I2 hljIjultlampAj nld-19H2) (0 3288 1111(ttIlL 144 66871 717 154 929 152 I Avergm LouUk KaLe of P(NJL LIR (7)I

(a) 19w--70 30 23 2t 26 17(b) 1970-62 27 194 Ptrcantap lka1P~Luijtjj I 26 29 18

InToul dljulachn(19J2) 71 76 88 765 13per iplta (doL1Lu 1982) 1 y0 200 00( V0 310 6 Creth of WP () 1

(a) I9w-70 b7 34 25 37 46(1) 1970-82 50 36 27 41 457 Agricultures Slave of Ag ttL6

12i (19812 ) I it 33 NA 47 278 Per Capita MP Trcm iu-lture

(1982 CUMtM dL]Lu) 1 88 69 NA 56 78 in (9 Cwh 3 fr AgrtriLrtue 1

() 1910J-70 49 19 NA 27 30(b) 1970-12 27 18 NA 23 32 10 Apiulturee Slave of lora[ laborForce (19W ) 1 57 71 93 74 5411Mci4sklibe Lx-ltWon(1981 X) I

(a) ltimry Camudi thL0 33 69(b) litules 4d Clouilg 1 32 65

23 24 56 1612 trdru1La j tpora X) I(1981

(a) luod 14 9 4 20 1913Ford Ai In (eteal(000 at) I

(a) 197475 19 1582 0 210 271(b) L98182 J68 41b 10 1076 19514axk of -xd Prokctjcn

(19-71-10) 145 135 112 121 17915 Average idexL FoodoCf Prvducrk

Per Capita 2 (I96-71-t13)

WtH12 101 83 94 15416 ImJi OtAgri-ctiral PrxkLELn 2 IUb

f196-71-0)0l91u 142 134 11 118 145

17Fuod Siqlly 7 (a) lurica per CnpiLa per dsy

(19111-4)Vegotabla Ixoducta lO055 1(1Y 1781 1810 2176Mli-l r hil 11)J I11) 111 G Ij

Mb Phrutdn (gram) Per LPtL

per Lly (1978-4 )) er b1 Prvducta 4L1 439 390 353 376Aiulal IlkwCte 141 46 67 53L 11Cccla-4 uf liI u 2

65

Per A9 IlLuralijrkr-r(91U)2 1591 1015 03hS 0356 076619Iercwawcp of Lexxi irrip (T) 2 704 233 99 177 24520 IerC ClIre of lawl Under erils 0-b534 613 96 118 39521Averul lld CL-erAlNJ(IIil914IT - 7l l j14 67-s 19f 242922 IlC L LILuI LI r 13 IHI | 13 163 b 1 65

J e lid ilIIi Jt ) s h 24 IrU 15 2 t l~d llg~ll~uldJ 2 ( 0 2UA 1U95 I U 2

26 lXc M= Li LubIlirr --Tmcor (9 7)7 -2 4)5 4481 2230 89

27 Ferrlizer Use pi krlai 8 (lMJk)l)hJ)--

S lO 91 6 1 770

Sourc 1 rU Blank orld[LaclTTmujt Ilu-rtI 1l9 2 F)19811FOd r1cluhtI Y -irtxk Vul 35 1982

1741N I AL l CExtiLL~er Yca d Kn 19 8t2Vo I J

1 ArailIa 1 Fernak2t cnu ir axuialcally acttvi Iatru1I InLlic ngrhultme iti c e miialy itiu ewI

oArea cy be pLat Irury thu LcL

Source ADB 1985

----------------------------------------------------------------------------------------------

----------------------------------------------------------------------------------------------

Table 1

REGRESSION RESULTS RELATING WHEAT AND RICE PRODUCTION TO RAINFALL

Dependent Constant Trend Sind Rain January Rain Auqust Rain DEC Rain R2 Variable DW ---------------------------------------------------------------------

WHEAT PRODUCTION (000) TONS

Pakistan -735397 37628 257 092

(33309) (1222) (231)

Pakistan -74223 0042 000029 0

(0003) (000013) (231)

Punjab -490921 25153 8377 093 (22472) (2651) (255)

Punjab -65462 0038 0001 093 (0003) (0003) (242)

Sind -99903 0054 000009 094 (0004) (00005) (18)

RICE PRODUCTION (000) TONS

Pakistan -118904 61482 063 (12722) (057)

Pakistan -3B574 0023 064 (0004) (057)

Punjab -14297 35402 -14297 071 (7255) (62427) (136)

Punjab -49503 0028 -0662 0005 089 (0003) (0138) (0001) (131)

Sind -22552 0015 031 (0006) (079)

fI The coefficient issignificantly different from zero at 01 probability level

41 Log function

Standard error

1

Appendix B 11w Agricultra ELtzN of Pi lefal in a R gkAul Ir6ptrtlv

0I)m-19142)

K1ak- Sri1lhJ L 101h II1al LiduJh L~aamp

I Lruxi Aiva (000 2) I2 hljIjultlampAj nld-19H2) (0 3288 1111(ttIlL 144 66871 717 154 929 152 I Avergm LouUk KaLe of P(NJL LIR (7)I

(a) 19w--70 30 23 2t 26 17(b) 1970-62 27 194 Ptrcantap lka1P~Luijtjj I 26 29 18

InToul dljulachn(19J2) 71 76 88 765 13per iplta (doL1Lu 1982) 1 y0 200 00( V0 310 6 Creth of WP () 1

(a) I9w-70 b7 34 25 37 46(1) 1970-82 50 36 27 41 457 Agricultures Slave of Ag ttL6

12i (19812 ) I it 33 NA 47 278 Per Capita MP Trcm iu-lture

(1982 CUMtM dL]Lu) 1 88 69 NA 56 78 in (9 Cwh 3 fr AgrtriLrtue 1

() 1910J-70 49 19 NA 27 30(b) 1970-12 27 18 NA 23 32 10 Apiulturee Slave of lora[ laborForce (19W ) 1 57 71 93 74 5411Mci4sklibe Lx-ltWon(1981 X) I

(a) ltimry Camudi thL0 33 69(b) litules 4d Clouilg 1 32 65

23 24 56 1612 trdru1La j tpora X) I(1981

(a) luod 14 9 4 20 1913Ford Ai In (eteal(000 at) I

(a) 197475 19 1582 0 210 271(b) L98182 J68 41b 10 1076 19514axk of -xd Prokctjcn

(19-71-10) 145 135 112 121 17915 Average idexL FoodoCf Prvducrk

Per Capita 2 (I96-71-t13)

WtH12 101 83 94 15416 ImJi OtAgri-ctiral PrxkLELn 2 IUb

f196-71-0)0l91u 142 134 11 118 145

17Fuod Siqlly 7 (a) lurica per CnpiLa per dsy

(19111-4)Vegotabla Ixoducta lO055 1(1Y 1781 1810 2176Mli-l r hil 11)J I11) 111 G Ij

Mb Phrutdn (gram) Per LPtL

per Lly (1978-4 )) er b1 Prvducta 4L1 439 390 353 376Aiulal IlkwCte 141 46 67 53L 11Cccla-4 uf liI u 2

65

Per A9 IlLuralijrkr-r(91U)2 1591 1015 03hS 0356 076619Iercwawcp of Lexxi irrip (T) 2 704 233 99 177 24520 IerC ClIre of lawl Under erils 0-b534 613 96 118 39521Averul lld CL-erAlNJ(IIil914IT - 7l l j14 67-s 19f 242922 IlC L LILuI LI r 13 IHI | 13 163 b 1 65

J e lid ilIIi Jt ) s h 24 IrU 15 2 t l~d llg~ll~uldJ 2 ( 0 2UA 1U95 I U 2

26 lXc M= Li LubIlirr --Tmcor (9 7)7 -2 4)5 4481 2230 89

27 Ferrlizer Use pi krlai 8 (lMJk)l)hJ)--

S lO 91 6 1 770

Sourc 1 rU Blank orld[LaclTTmujt Ilu-rtI 1l9 2 F)19811FOd r1cluhtI Y -irtxk Vul 35 1982

1741N I AL l CExtiLL~er Yca d Kn 19 8t2Vo I J

1 ArailIa 1 Fernak2t cnu ir axuialcally acttvi Iatru1I InLlic ngrhultme iti c e miialy itiu ewI

oArea cy be pLat Irury thu LcL

Source ADB 1985

Appendix B 11w Agricultra ELtzN of Pi lefal in a R gkAul Ir6ptrtlv

0I)m-19142)

K1ak- Sri1lhJ L 101h II1al LiduJh L~aamp

I Lruxi Aiva (000 2) I2 hljIjultlampAj nld-19H2) (0 3288 1111(ttIlL 144 66871 717 154 929 152 I Avergm LouUk KaLe of P(NJL LIR (7)I

(a) 19w--70 30 23 2t 26 17(b) 1970-62 27 194 Ptrcantap lka1P~Luijtjj I 26 29 18

InToul dljulachn(19J2) 71 76 88 765 13per iplta (doL1Lu 1982) 1 y0 200 00( V0 310 6 Creth of WP () 1

(a) I9w-70 b7 34 25 37 46(1) 1970-82 50 36 27 41 457 Agricultures Slave of Ag ttL6

12i (19812 ) I it 33 NA 47 278 Per Capita MP Trcm iu-lture

(1982 CUMtM dL]Lu) 1 88 69 NA 56 78 in (9 Cwh 3 fr AgrtriLrtue 1

() 1910J-70 49 19 NA 27 30(b) 1970-12 27 18 NA 23 32 10 Apiulturee Slave of lora[ laborForce (19W ) 1 57 71 93 74 5411Mci4sklibe Lx-ltWon(1981 X) I

(a) ltimry Camudi thL0 33 69(b) litules 4d Clouilg 1 32 65

23 24 56 1612 trdru1La j tpora X) I(1981

(a) luod 14 9 4 20 1913Ford Ai In (eteal(000 at) I

(a) 197475 19 1582 0 210 271(b) L98182 J68 41b 10 1076 19514axk of -xd Prokctjcn

(19-71-10) 145 135 112 121 17915 Average idexL FoodoCf Prvducrk

Per Capita 2 (I96-71-t13)

WtH12 101 83 94 15416 ImJi OtAgri-ctiral PrxkLELn 2 IUb

f196-71-0)0l91u 142 134 11 118 145

17Fuod Siqlly 7 (a) lurica per CnpiLa per dsy

(19111-4)Vegotabla Ixoducta lO055 1(1Y 1781 1810 2176Mli-l r hil 11)J I11) 111 G Ij

Mb Phrutdn (gram) Per LPtL

per Lly (1978-4 )) er b1 Prvducta 4L1 439 390 353 376Aiulal IlkwCte 141 46 67 53L 11Cccla-4 uf liI u 2

65

Per A9 IlLuralijrkr-r(91U)2 1591 1015 03hS 0356 076619Iercwawcp of Lexxi irrip (T) 2 704 233 99 177 24520 IerC ClIre of lawl Under erils 0-b534 613 96 118 39521Averul lld CL-erAlNJ(IIil914IT - 7l l j14 67-s 19f 242922 IlC L LILuI LI r 13 IHI | 13 163 b 1 65

J e lid ilIIi Jt ) s h 24 IrU 15 2 t l~d llg~ll~uldJ 2 ( 0 2UA 1U95 I U 2

26 lXc M= Li LubIlirr --Tmcor (9 7)7 -2 4)5 4481 2230 89

27 Ferrlizer Use pi krlai 8 (lMJk)l)hJ)--

S lO 91 6 1 770

Sourc 1 rU Blank orld[LaclTTmujt Ilu-rtI 1l9 2 F)19811FOd r1cluhtI Y -irtxk Vul 35 1982

1741N I AL l CExtiLL~er Yca d Kn 19 8t2Vo I J

1 ArailIa 1 Fernak2t cnu ir axuialcally acttvi Iatru1I InLlic ngrhultme iti c e miialy itiu ewI

oArea cy be pLat Irury thu LcL

Source ADB 1985

Append ix C ~Data Shetf

7 f7

155 774 H1 1320 132 13v 1 1 q ~ Iv

197i

198i

IiK 1964

1985 1926

8tiI1 94

198 8367 995 M9005 11475

192 134 h2414 1682

11703 13000

al 69763

~ 03 66K7 t24 89i 7-2

7398 7343

72-5 7406

1422w 1431 1316 1455 15hB 1e 15o5 1676 1482 1612 1757

54i -572 141224 166i 646 143ff] 1691I 63i 13156 151t) b75 14880 1678 824 1567i 1754 87-1843NO 164 0 iss~ 1764 82 16780) 45~ 14E20 1671) 7P0

16124 1828 791) l75J8

445v 45i4 4811 4861 520i 5454 546t 56is 5760 575

5660

07 v7

Vto 077i (176~ 1)7q 7a8 7

(78 0- ~ vu

271 2617 2737 255C 3272 -121

12 54 3C 44 v

3315 2c70

443~ 5 1565k 1553 1615 151] 1a1 173 1 74 2 17

165i 1588

160b4

1710 174 18Y9

2026 24 1 H768 191 98

1996 1870

lF A C i 4SA b ~LE E FE E h h 1

3 C241

LI

2231

FERTi

3247

FR i E WIL

Y v

Iu242

5

3 5 k2 c 4 4 i2153e 11

iS 5

t i 2 3 C i

v2 0 C 2-

C

-C

212 2412

224

33 42

3 251 1i3355

325

53ii 54 45

4 5

52 -

774

21 ) 123C355

35

251 j 3 3553 0(

51

5

407 b3

424 5

7 b 7

I04585-0

)85

1) vo

Z3

- 1 39 45

(1

7

72s 4

7

71

75amp

75-

123Y4

v7

123 35t H 81 0t5 u) 074 744 70 2t 48 1j 3 Il 3 286 aI 9 074 7b3070 43 38 41 118 334 28k 307 850 074 07o 077 7b7 84 29 jb (4 279 284 590 077 7 687 89 37 37 37 111 41F 3717 17 90 077 087 091 914 12b 72 9i 149 245 14 377 o70 07 091 084 837 198 107 153 737 254 185 164 73v 0i 084 100 995 122 17 70

137 211I 55 185 730 i 64 10 10 1066 182 b2 22 2b 192 i 5 5 94 If( (L 115 114 75 15 i

S1o 3 8 5 113 i21 (5 I - br 1I 24 2i 87 I6

I5 174 1 o 12T 217L( i17L v ot 72 159 248 1 I~ ~ U 1~ 2 3k 17 34ci l 3 i4 15 20 9 o( 117 23 93 i55 8 57 I 30 Iv 780 I3 1

s160 L7 vo

TI RR 1( CANihR TWAT FETIR TNELL OATER4 RPRO(b NTEF WFROD 13 13 7u 22 13 45 48 8 25 2F4 13 13 92 711 282 14b 4o 491- 25 2750v

131 131 94 612 311 158 52 448 3i) 248(17 13 133

13b 133

9o 97

8t1 77

34) 319

17 16

49 5)1) 5(1 462

31 27

24oi3 28422

136 113 9 a5 479 194 5f 523 34 255 2 138 138 97 5 4o2 205 42 57 3b 254c 142 142 I1 8 4 478 21t 53 55 377 2 c 145 145 6a874 51 22 51 b41b 2 147 149

147 14

l04 f7

4 92

72 o5

24 25

5 5

574 52 93

3

8 2-9423

15 5 15 1 78 2i3 5 5 39 42 1533 153 11 24 2 54755 43 25

im153 11 Is 817 27 bi 5434 44 2b5

lF A C i 4SA b ~LE E FE E h h 1

3 C241

LI

2231

FERTi

3247

FR i E WIL

Y v

Iu242

5

3 5 k2 c 4 4 i2153e 11

iS 5

t i 2 3 C i

v2 0 C 2-

C

-C

212 2412

224

33 42

3 251 1i3355

325

53ii 54 45

4 5

52 -

774

21 ) 123C355

35

251 j 3 3553 0(

51

5

407 b3

424 5

7 b 7

I04585-0

)85

1) vo

Z3

- 1 39 45

(1

7

72s 4

7

71

75amp

75-

123Y4

v7

123 35t H 81 0t5 u) 074 744 70 2t 48 1j 3 Il 3 286 aI 9 074 7b3070 43 38 41 118 334 28k 307 850 074 07o 077 7b7 84 29 jb (4 279 284 590 077 7 687 89 37 37 37 111 41F 3717 17 90 077 087 091 914 12b 72 9i 149 245 14 377 o70 07 091 084 837 198 107 153 737 254 185 164 73v 0i 084 100 995 122 17 70

137 211I 55 185 730 i 64 10 10 1066 182 b2 22 2b 192 i 5 5 94 If( (L 115 114 75 15 i

S1o 3 8 5 113 i21 (5 I - br 1I 24 2i 87 I6

I5 174 1 o 12T 217L( i17L v ot 72 159 248 1 I~ ~ U 1~ 2 3k 17 34ci l 3 i4 15 20 9 o( 117 23 93 i55 8 57 I 30 Iv 780 I3 1

s160 L7 vo

TI RR 1( CANihR TWAT FETIR TNELL OATER4 RPRO(b NTEF WFROD 13 13 7u 22 13 45 48 8 25 2F4 13 13 92 711 282 14b 4o 491- 25 2750v

131 131 94 612 311 158 52 448 3i) 248(17 13 133

13b 133

9o 97

8t1 77

34) 319

17 16

49 5)1) 5(1 462

31 27

24oi3 28422

136 113 9 a5 479 194 5f 523 34 255 2 138 138 97 5 4o2 205 42 57 3b 254c 142 142 I1 8 4 478 21t 53 55 377 2 c 145 145 6a874 51 22 51 b41b 2 147 149

147 14

l04 f7

4 92

72 o5

24 25

5 5

574 52 93

3

8 2-9423

15 5 15 1 78 2i3 5 5 39 42 1533 153 11 24 2 54755 43 25

im153 11 Is 817 27 bi 5434 44 2b5

123 35t H 81 0t5 u) 074 744 70 2t 48 1j 3 Il 3 286 aI 9 074 7b3070 43 38 41 118 334 28k 307 850 074 07o 077 7b7 84 29 jb (4 279 284 590 077 7 687 89 37 37 37 111 41F 3717 17 90 077 087 091 914 12b 72 9i 149 245 14 377 o70 07 091 084 837 198 107 153 737 254 185 164 73v 0i 084 100 995 122 17 70

137 211I 55 185 730 i 64 10 10 1066 182 b2 22 2b 192 i 5 5 94 If( (L 115 114 75 15 i

S1o 3 8 5 113 i21 (5 I - br 1I 24 2i 87 I6

I5 174 1 o 12T 217L( i17L v ot 72 159 248 1 I~ ~ U 1~ 2 3k 17 34ci l 3 i4 15 20 9 o( 117 23 93 i55 8 57 I 30 Iv 780 I3 1

s160 L7 vo

TI RR 1( CANihR TWAT FETIR TNELL OATER4 RPRO(b NTEF WFROD 13 13 7u 22 13 45 48 8 25 2F4 13 13 92 711 282 14b 4o 491- 25 2750v

131 131 94 612 311 158 52 448 3i) 248(17 13 133

13b 133

9o 97

8t1 77

34) 319

17 16

49 5)1) 5(1 462

31 27

24oi3 28422

136 113 9 a5 479 194 5f 523 34 255 2 138 138 97 5 4o2 205 42 57 3b 254c 142 142 I1 8 4 478 21t 53 55 377 2 c 145 145 6a874 51 22 51 b41b 2 147 149

147 14

l04 f7

4 92

72 o5

24 25

5 5

574 52 93

3

8 2-9423

15 5 15 1 78 2i3 5 5 39 42 1533 153 11 24 2 54755 43 25

im153 11 Is 817 27 bi 5434 44 2b5

TI RR 1( CANihR TWAT FETIR TNELL OATER4 RPRO(b NTEF WFROD 13 13 7u 22 13 45 48 8 25 2F4 13 13 92 711 282 14b 4o 491- 25 2750v

131 131 94 612 311 158 52 448 3i) 248(17 13 133

13b 133

9o 97

8t1 77

34) 319

17 16

49 5)1) 5(1 462

31 27

24oi3 28422

136 113 9 a5 479 194 5f 523 34 255 2 138 138 97 5 4o2 205 42 57 3b 254c 142 142 I1 8 4 478 21t 53 55 377 2 c 145 145 6a874 51 22 51 b41b 2 147 149

147 14

l04 f7

4 92

72 o5

24 25

5 5

574 52 93

3

8 2-9423

15 5 15 1 78 2i3 5 5 39 42 1533 153 11 24 2 54755 43 25

im153 11 Is 817 27 bi 5434 44 2b5