Agribusiness Review - Vol. 12 - 2004

Paper 5
ISSN 1442-6951

AGRICULTURAL PROCESSING AND THE WA ECONOMY: A GENERAL EQUILIBRIUM ANALYSIS

Peter Johnson and Nazrul Islam [1]

Peter Johnson, Economic Research Centre, University of Western Australia .
Nazrul Islam, Department of Agriculture, Western Australia .

ABSTRACT

This paper investigates the impact of an expansion in agricultural processing on the Western Australian economy by modifying and applying a Computable General Equilibrium (CGE) economic model of Western Australia (called WAM).  WAM was used to simulate the effects of a $1 million expansion in eight agricultural processing industries.  The results show that there is a range of positive impacts from agricultural processing.  On average, a $1 million expansion in agricultural processing is estimated to increase the State's GSP (Gross State Product) by $649,000, and total output by $1.9 million.  The expansion of the Wine and spirits industry is estimated to have the largest impact while the Textile fibres, yarns and woven fabrics industry has the smallest impact on the Western Australian economy.

1. Introduction

With its favourable factor endowments, Western Australia enjoys a comparative advantage in agricultural production and export.  The State produces a wide range of export oriented agricultural commodities, including, broadacre crops (predominantly wheat), wool, sheep, cattle and other livestock.  In 1998/99, the gross value of agricultural production in WA stood at $4.9 billion [2] , which represents about 15 per cent of national production.  During the past two decades, the agricultural sector in WA grew at an average rate of over 6 per cent per annum  (Islam, 2000).  However, although WA is a major producer of agricultural commodities, and has a wealth of natural advantages including a clean environment and a stable and strong economy, not much agriculture-based processing has taken place in the State.  This is in spite of the fact that for a long time an important policy objective of the WA government has been to expand the local processing of primary products before export.  This policy is in place because it is believed that downstream processing is important for ensuring the continued growth of WA agriculture.

While accounting for around 15 per cent of Australia 's primary agricultural output, WA produces only about 7 per cent of the gross product of the national food manufacturing industry (ABS, 2001a and 2001b).  So, while about 75 per cent of WA's agricultural output is exported, it is mostly in unprocessed form.  Between 1995 and 1999, on average, only about 12 per cent of the total WA agricultural exports were in processed form.  By comparison, over 50 per cent of the agricultural exports from the rest of Australia were in processed form.  For some individual commodities, the lack of processing in WA is even worse.  For example, WA accounts for only 4 per cent of the national exports of meat products, while its share in national live animal exports is over 40 per cent.  Australia as a whole lags behind other exporters of agricultural processed commodities [3] and WA clearly lags behind the rest of Australia in agricultural processing activities. 

Given the marked differences between the prices of processed agricultural products and unprocessed agricultural commodities, one might suspect that the WA economy is losing heavily by not processing its primary products before export.  With market access improving (due to multilateral trade negotiations under the auspices of GATT/WTO and APEC) and with growing demand for processed foods, the prospect for downstream processing of primary products in WA has improved.  At the federal level, the government has adopted a number of programs and initiatives to improve the international competitiveness and export orientation of the agricultural processing industries (see, e.g., National Food Industry Strategy report, AFFA, 2002).  With WA's low level of agricultural processing, the State is failing to take advantage of these opportunities.

To appreciate the contribution that expanded agricultural industries may have on the WA economy, in this paper we simulate the impact of a $1 million expansion in a variety of agricultural processing industries.  This is accomplished using a Computable General Equilibrium (CGE) model of the Western Australian economy.  This model helps us to obtain answers to the following questions:

• To what extent do primary agricultural and other non-agricultural industries get affected, via inter-sectoral linkages, due to an expansion in WA agricultural processing industries?
• By how much would income and employment opportunities change if the State's agricultural processing industries expanded?

The remainder of this paper is divided into four sections. Section 2 provides an overview of the agricultural and agricultural processing industries in WA.  In Section 3, the characteristics of the CGE model for the WA economy is described, while in Section 4, the model is applied and its results are discussed in detail.  Finally, Section 5 concludes the paper and presents a summary of the major findings.

2.         FOOD AND AGRICULTURAL PROCESSING IN WESTERN AUSTRALIA

Although the agriculture sector is relatively important to the WA economy, contributing more than four per cent to the State's GSP [compared to less than three per cent for the rest of Australia (ROA) (ABS, 2001)], the State's food processing sector accounts for a little more that one per cent of the State's GSP, as compared to about three per cent for the ROA (Islam and Johnson, 2003). [4]

The relative lack of food processing in WA is in part a reflection of the State's relatively low share of Australian manufacturing.  As can be seen in Table 2.1, WA's share of the national food manufacturing value-added in only 6.5 per cent (see column 3), while its share of total manufacturing value-added is only 7.4 per cent (see column 5).  However, WA also ranks the lowest amongst the Australian States in terms of food manufacturing share (18 per cent) of total manufacturing (see column 6 of Table 2.2), indicating that the lack of food processing in the State is due to more than just WA's relatively small manufacturing base.

Table 2.1.         Food 1 manufacturing value added in Australian States, 1999/2000

Notes:          Notes:

1. Processed foods including beverages and tobacco.
2. Value added is Gross Domestic Product equivalent.
3. Entries in column 2 as percentage of the corresponding entries in column 4.

Source:  ABS (2001a and 2001b)

A detailed look at the extent to which WA agricultural commodities are processed and exported is presented in Table 2.2.  As can be seen in row 7, only 25 per cent (14 + 11 per cent, see columns 3 and 5) of the State's primary agricultural commodities are processed in some form or other.  The remaining 75 per cent is marketed in raw commodity form.  The situation is even more disappointing for the major commodity groups such as cereals, pulses and oilseeds and wool.  These commodities comprise about 70 per cent of the State's gross value of agricultural production (GVAP) (Islam, 2000).  Cereals (mainly wheat) comprises about 45 per cent of the GVAP but only four per cent are processed, including two per cent exports.  On the other hand, the meat industry processes about 80 per cent (including 25 per cent for exports) of meat producing animals in WA.  However, recent trends indicate that increasing proportions of beef cattle are now exported live (Islam and Johnson, 2003).  As mentioned earlier, WA accounts for only four per cent of the national exports of meat products, while its share in national live animal exports is 40 per cent.

TABLE 2.2        Percentage distribution of processed and unprocessed agricultural commodities produced in WA for domestic use and exports in a typical year

Notes:         

(a) Refers to cattle and sheep stocks and (b) Refers to the dairy industry, the unprocessed amount of milk refers to white market milk.  Technically, all market milk also goes through some form of processing, bottling and packaging.

Source:  Islam (1997)

Table 2.2 reveals that an insignificant proportion of agricultural commodities produced in Western Australia are processed and exported.  Overall, although around 75 per cent of the primary production is exported, only 11 per cent is in processed form.  This indicates that there are tremendous opportunities to benefit from the expansion of processing primary agricultural commodities in WA.

In this paper we take the position that untapped opportunities exist in Western Australia for the further development of the State's agricultural processing sectors.  Islam and Johnson (2003) identified impediments to expansions of agricultural processing in Western Australia .  On the basis that these impediments can be reduced, through a combination of State Government policy and industry innovation, there exists within the State opportunities to process existing agricultural produce without the need for any expansion to local or world demand.

3. The WA Model

The use of Computable General Equilibrium (CGE) models for economic analysis began in Australia with the creation of the ORANI model (Dixon et al. 1982).  ORANI, in its original form, is a single-region model of the Australian economy; that is, it models the entire Australian economy, without any consideration of state level activities.  Since the inception of ORANI, a variety of CGE models have been developed in Australia , including models which capture state level activities.  One such model is WAM (the WA model) (Clements et al . 1996) which is used for the analysis in this report.

3.1        Characteristics of CGE Models

CGE models have many advantages over other methods of economic analysis, such as input-output analysis.  Whereas input-output analysis assumes the economy remains static (i.e.  that price levels, labour to capital ratios and import shares remain unchanged throughout the analysis), CGE models are able to incorporate and predict changes to the economic structure.  CGE models are able to do this because they contain equations describing a wide range of economic activities, including production, consumption, investment, employment, taxation and trade.

CGE models consist of two major components:  the equations and the database.  While the equations give the model its predictive power, they are of no use without a comprehensive data set.  The data incorporated into the model specifies the structure of the economy being analysed, and tells the model how variables react to changes in other variables.  The economic structure is specified in CGE models with the inclusion of an input-output table.  Input-output tables describe the transactions occurring within the economy in great detail, including, the transactions occurring between industries and the transactions occurring between industries and final consumers.  How variables react to each other is specified by the elasticities of the database.

3.2        The WA Model

The WA model (WAM), used for the analysis in this report, is similar in many respects to ORANI.  Just like ORANI, WAM is formulated in percentage change terms.  WAM also treats Western Australia as a single region, and contains an extensive set of equations describing production, consumption, investment, employment, taxation and trade within the State's economy.  Therefore, it can be said that WAM is structured in a fairly standard way for CGE models in Australia .  What distinguishes WAM, and makes it such a useful tool for economic analysis in Western Australia , is the model's database.  The WAM database contains the most detailed information available on the economy of Western Australia .  The input-output table currently used in WAM is a 108-sector table for the financial year 1994/95.  The table is based on the 105-sector table for WA developed by Johnson (2001), with additional detail provided in primary agricultural industries (see Appendix 1).

The original version of WAM (Clements et al. 1996) contained less detail than the current version, as its database was based on the 42-sector input-output table for 1989/90 (Clements and Ye, 1995).  Even though it was less disaggregated than the current version of the model, it was still a highly effective tool for economic analysis, and was used to analyse such issues as:

• the impact of new mining and minerals processing projects on the economy of Western Australia (Clements et al. 1996).
• the impact of increased minerals production on the economy of Western Australia (Ahammad and Clements, 1999),
• the impact of minerals industry growth on employment in different regions of Western Australia (Clements and Johnson, 2000),
• the impact of tariffs on the Western Australian economy (Ahammad and Greig, 2000), and
• the impact of lower energy costs on the Western Australian economy (Clements et al. 2002, Chapter 3).

WAM also became the basis for a variety of more specialised models:  models such as WAT  -  a two-regional model of the WA economy  -  which was used to determine the impact of the Hot Briquetted Iron plant on the economy of the Pilbara region (Johnson, 1999), and WAE  -  a CGE model that incorporates energy substitution  -  which was used to investigate the impact of greenhouse gas reduction policies on the WA economy (Ahammad et al. 2001).

3.3        Modifications to WAM

In WAM, there are only two primary factors of production, labour and capital  -  where capital, in agricultural sectors, is a composite of land and capital.  It is assumed in WAM simulations that labour is mobile across industries, and that the total supply of labour is not limited.  Therefore, all industries can demand as much or as little labour as they require.  Capital, on the other hand, is assumed to be industry specific and fixed in supply.  Now, for certain primary agricultural industries this treatment of capital is unnecessarily, and unrealistically restrictive.  In the application of WAM in this paper, we assume that some agricultural industries can 'share/swap' capital.  The industries covered by this assumption are separated into two groups:

• Group A: Sheep meat (1), Wool (2), Cereals (3) and Pulses and oilseeds (4); and
• Group B: Horticulture (8), New industries (9) and Dairy cattle (10).

The numbers after each industry represent their position within WAM's industry structure. 

For the industries within each group, the capital stocks are allowed to vary; however, the capital stock for the group as a whole is assumed to be fixed, so that the following equations hold:

(3.1)      K_A = K₁ + K₂ + K₃ +K₄ , and

(3.2)      K_B = K₈ + K₉ + K₁₀,

where    K_i (i = 1-4, 8-10) represents the capital stock in each industry, and K_A and K_B are both fixed.

As part of WAM's determination of economic variables, the change in the price paid to units of capital is calculated.  This price, P_i^K   (where i = 1-4 for Group A industries, and i = 8-10 for Group B industries), provides the signal for capital redistribution within each group.  For example, if the price paid to capital in the Sheep meat industry  ( P₁^K)  exceeds the price paid to Cereals ( P₃^K ),  then capital will shift from the Cereals industry to the Sheep meat industry until the prices are equal.  In other words, capital stocks redistribute between industries in Group A until

(3.3)      P₁^K = P₂^K = P₃^K = P₄^K.

Similarly, for Group B industries capital redistribution occurs until

(3.4)      P₈^K = P₉^K = P₁₀^K.

Equations (3.1) to (3.4) are in levels, while, as stated previously, WAM is formulated in percentage changes.  The percentage change versions of these equations are not presented here; however, they are contained in Appendix 2.  Appendix 2 also contains an alternative approach for deriving the percentage change versions of equations (3.3) and (3.4).

3.4        Impact of the modifications

With the modifications described above, there is, potentially, a significant effect on model outcomes for those industries in Groups A and B.  To describe the nature of these effects, we present a simple graphical analysis using production possibility frontiers.  To do this, we assume the existence of an economy which produces only two goods, A and B.  Panel 1 of Figure 3.1, presents the production possibility frontier for these two goods.  The quantity of good A produced ( Q_A) is shown on the vertical axis, while the quantity of good B produced ( Q_B) is shown on the horizontal axis.  The curve shown in panel 1 is the production possibility frontier for the production of these two goods, under the assumption that the capital employed in this two-good economy is industry specific, and cannot be shifted from the production of A to the production of B, and vice versa.  In this simple system, the point at which production occurs is the point where the slope of the production possibility frontier is equal to the slope of the price line; where the slope is given by the price of good B ( P_B ) relative to the price of good A ( P_A ).

Initially, with the relative price at P_B/P_A, the economy produces at point x on the production possibility frontier  -  which we assume to be a position of long-run stability, where capital in each industry is employed at maximum efficiency.  Next, due to some disturbance in the economy, prices shift to P'_A and P'_B (relative price P'_A/P'_B ), and a new equilibrium is established at the point y, where the production of good A has diminished, and the production of good B has increased.

Figure 3.1        Production possibilities under different capital assumptions.

Now, consider panel 2 of Figure 3.1.  Here, it is assumed that capital is not industry specific, but may be shifted between industries.  The original production possibility frontier is shown as the dotted curve in panel 2, with the new frontier shown as the solid curve.  Note that the new curve touches the old at only one point: x.  Recall that it was stated above, that point x represented a position of long-run stability, where capital in each industry is employed at maximum efficiency; therefore, no additional production of A or B is available at point x by redistributing capital.  The remainder of the new production possibility frontier is outside the old frontier, and is the envelope of all possible capital-constrained production possibilities.

Given the same economic disturbance, and the same shift in prices, that we saw in panel 1, a new equilibrium is established at point y¹ in Panel 2. As is clear, the movement from point x to point y¹ represents a more dramatic shift in the production pattern than does the movement from x to y, i.e. there is a greater reduction in the production of good A,  and a greater increase in the production of good B.  These larger changes occur because of the ability of capital to shift between the two industries.  This analysis suggests that within WAM, under the assumption of joint capital, it can be expected that more pronounced changes in production will occur within Group A and Group B industries than could otherwise be expected.  The magnitude of these effects will be studied in Section 4.

3.5        The simulations

The industry structure used in WAM includes ten primary agricultural industries.  Also within WAM's industry structure are numerous industries that process the output of these primary agricultural sectors.  These include:

• Meat and meat products
• Dairy products
• Fruit and vegetable products
• Oils and fats
• Flour mill products and cereal foods
• Beer and malt
• Wine and spirits; and
• Textile fibres, yarns and woven fabrics.

As discussed in Section 2, in this paper we take the position that opportunities exist for expansion of agricultural processing in Western Australia, and that these opportunities could be exploited without the need for any further increase in local or world demand (although changes to State Government policy and local industry innovation may be required).  In Section 4, we use WAM to estimate the impact on the economy of Western Australia of an expansion in the above eight processing sectors.  In other words, we investigate the impact of positive supply side shocks to these industries.  So as to provide easily comparable results, the simulations are performed on the basis of a $1 million expansion in the output of each of these industries.  In order to conduct these simulations, the $1 million expansions were first converted into percentage changes in the output of these industries.  These changes then provide the inputs or 'shocks' to the model.  The calculation of these shocks is presented in Appendix 3.

4.         SIMULATION RESULTS

In this section, we present the results of the simulations designed to predict the impact on the WA economy of a $1 million expansion in each of eight agricultural processing industries. [5]   The simulations were performed using the WA model (WAM) described in the previous section.  We begin by looking at the impact of the expansion on key macroeconomic variables, before considering industry level impacts.

4.1        Macroeconomic impacts

Consider the results presented in Table 4.1.  For the $1 million increase in the output of the eight agricultural processing industries shown in column 1, the resulting increases in real Gross State Product (GSP), the consumer price index (CPI), employment, imports and exports, are provided in columns 2 to 6 of the table.  Clearly, the table shows that the agricultural processing industry with the most beneficial impact on the State's GSP is the Wine and spirits industry, with GSP estimated to grow by $1,035,000 for every $1 million increase in its output.  Beer and malt is the next most expansionary agricultural sector, followed by Fruit and vegetable products.  Textile fibres, yarns and woven fabrics, with a GSP impact of $381,000, has the lowest impact.  To understand the ranking of real GSP impacts revealed in Table 4.1, consider Table 4.2, which shows the input coefficients for the eight agricultural processing industries.  Row 26 of Table 4.2 shows the direct contribution to GSP of a $1 expansion in the output of each industry.  Beer and malt, Fruit and vegetable products and Wine and spirits have, in that order, the highest direct GSP contributions of the eight industries shown.  Textile fibres, yarns and woven fabrics has the lowest direct contribution.  The ordering of direct GSP contribution is therefore very similar to the ordering of impacts shown in Table 4.1.  This not surprising as the results in Table 4.1 show the direct plus indirect effects, or total effects.  However, the direct GSP contribution alone does not account for the ranking of real GSP impact shown in Table 4.1, however they are critical.  The direct effects are important for two main reasons.  The first, and obvious one, is that they show the direct impact of the expansion on real GSP.  The second reason is that private consumption in WAM is a linear function of real GSP.  The direct stimulus from the expansion in agricultural processing flows immediately on to consumer spending.  Why then is the direct GSP effect not a perfect predictor of the real GSP effects shown in Table 4.1?  The reason is the flow-on effects captured by the WAM simulations.  Row 20 of Table 4.2 shows, for each agricultural processing industry, the share of inputs to production sourced from within Western Australia .  Consider the industries Beer and malt, Fruit and vegetable products and Wine and spirits (columns 7, 4 and 8 respectively).  Wine and spirits has the highest local input use of these three industries, indicating that the flow-on effects generated from its expansion are likely to be higher than the flow-on effects generated by the other two industries (although this should be viewed as a 'rule-of-thumb' rather than a rigorous relationship).  This is why Wine and spirits ranked the highest in real GSP impact, even though it ranks only third in direct GSP impact.

Table 4.1 shows that the CPI and employment impacts of agricultural expansion follow a similar pattern to the total GSP impact, with the $1 million expansion in Wine and spirits creating the most jobs, 22, and increasing the CPI by 0.0015 per cent  -  this CPI increase is rather insignificant, but remember we are dealing with a relatively small increase in output.  The expansion in the Textile fibres, yarns and woven fabrics industry increases employment by only 11 persons, and increases the CPI by 0.0005 per cent.  The CPI effects follow precisely the same ranking as the impacts on real GSP.  This is to be expected.  As described above, private consumption in WAM is a linear function of real GSP; therefore, it is to be expected that the industries generating the highest GSP effects will increase private consumption by the most, and hence produce the highest increase in consumer prices.

Table 4.1.         Macroeconomic impact of an expansion in agricultural processing industries

Consider the impact on imports, shown in column 5 of Table 4.1.  The expansion of the Textile fibres, yarns and woven fabrics industry produces the smallest increase in imports.  The Oils and fats industry produces the largest increase.  This is not a surprising result, as the Textile fibres, yarns and woven fabrics industry has one of the lowest import propensities among the agricultural processing industries (just over three per cent), while oils and fats has the highest (22 per cent), as can be seen from row 24 of Table 4.2.

Next, column 6 of Table 4.1 shows the increase in exports resulting from the expansion in agricultural processing.  The smallest increase in exports occurs with the expansion of the Textile fibres, yarns and woven fabrics industry, and the largest occurs with the Oils and fats industry.  This is the same result as we found for imports, which is to be expected.  Industries that consume few imports will consume more locally produced commodities when they expand.  Much of this increased domestic consumption will be at the expense of exports.  So, while most of the expanding industries output may be exported, there will be a high level of absorption by that industry of local commodities that would otherwise have been exported.  Likewise, high importing industries have lower domestic absorption, and consequently their expansion results in higher exports.

4.2        Industry impacts

In addition to its ability to estimate impacts at an economy wide level  -  the macroeconomic effects  -  WAM is able to estimate impacts for each of the 108 industries in the model.  Here, we consider these industry level impacts.  However, before examining the results of the WAM simulations, it is useful to discuss the industry-industry interactions in the model's input-output database, as the relationships revealed will help us to interpret the modelling results.

Table 4.2          Input coefficients for agricultural processing industries (percentages)

Table 4.2 contains a summary of the key industry relationships from the input-output table used in the WAM database.  The columns of the table present the consumption shares (in percentages) for intermediate inputs and primary factors in agricultural processing industries.  For example, column 2 summarises the purchases made by the Meat and meat products industry when producing its output.  To save space, consumption from all of the 108 industries in the database is not provided.  What is provided is a full list of the input shares of the primary agricultural industries (rows 1 to 10 of Table 4.2), a full list of input shares from the eight agricultural processing industries (rows 11 to 18), the total share of inputs of other - non-agricultural - goods and services (row 19), the share of total intermediate inputs in production (row 20), and finally (in rows 21 to 24), the share of inputs/costs covered by Compensation of employees (wages), Gross operating surplus (profits), Taxes and Imports.  As the figures in each column represent cost/input shares in percentage terms, they sum to one hundred, as shown in row 25.

From the information in Table 4.2 we can see which industries  -  particularly which primary agricultural industries  -  are most closely associated to the eight agricultural processing industries.  Starting with the Meat and meat products industry (column 2 of Table 4.2), we see that the industry takes inputs from the Sheep meat (12 per cent), Beef cattle (28 per cent), Pigs (5 per cent) and Poultry (9 per cent) sectors, all of which will benefit from any expansion in the output of Meat and meat products.  The expansion of the Dairy products industry (column 3) will be of most benefit to the Dairy cattle industry, as Dairy cattle supplies 34 per cent of its inputs.  An expansion in the Fruit and vegetable products industry (column 4) will benefit Horticulture (with 3 per cent of inputs) and New industries (5 per cent) the most.  The Pulses and oilseeds industry (with 4 per cent of total inputs) is the most significantly linked primary agriculture sector to the Oils and fats industry (column 5).  In spite of this, it is interesting to note that Oils and fats gains an even higher share of its inputs from the Meat and meat products industry (5 per cent), with an even larger share still supplied from within the industry itself (10 per cent).  Flour mill products and cereal foods (column 6) derives 18 per cent of total inputs from Cereals, while the Beer and malt industry (column 7) derives 10 per cent of its inputs from Cereals.  The Wine and spirits industry (column 8) takes significant inputs from Horticulture (5 per cent) and New industries (11 per cent).  Finally, the Textile fibres, yarns and woven fabrics industry (column 9) derives a massive 58 per cent of its total inputs from Wool, clearly the most significant relationship demonstrated in Table 4.2.

Keeping the relationships between the agricultural processing and primary agriculture industries in mind will aid with the interpretation of the WAM simulation results presented in Table 4.3.  The impact of the expansion of the agricultural processing industries on the primary agricultural sectors are shown in rows 1 to 10 of the table.  Consider first the results for the expansion of the Meat and meat products industry (column 2).  As expected, we see an expansion in the primary agricultural industries of Sheep meat, Beef cattle, Pigs and Poultry, although the expansion in the later three sectors is relatively small compared to the expansion in Sheep meat output of $134,000.  Recall that the industry Sheep meat is part of a group of agricultural industries which share capital (the Group A industries described in the previous section).  These industries are capable of shifting capital (which includes agricultural land) between the production of the different Group A commodities (Sheep meat, Wool, Cereals, and Pulses and oilseeds) even though the total stock of capital available has not changed.  With the expansion of the Meat and meat products industry, the demand for Sheep meat, Beef

Table 4.3.         Industry impact of an expansion in agricultural processing industries ($'000)

cattle, Pigs and Poultry all increase.  For the industries Beef cattle, Pigs and Poultry, most of the increased domestic demand for their output is met by reducing exports, with only a small increase in their total production.  Sheep meat, which is able to gain access to more capital  -  at the expense of the Wool, Cereals and Pulses and oilseeds sectors  -  is able to meet more of the increased domestic demand by increasing production.

With the ability of the Sheep meat industry to command more capital at the expense of the Group A industries, it is not surprising to see that the output of these other industries diminishes, with the output from the Cereals industry falling by $66,000.  It is interesting to note that the output of the Wool industry falls by a far less significant $8,000.  This indicates that farmers will not increase Sheep meat production by significantly shifting capital away from Wool, but, rather, by decreasing the capital (which we should remember includes land) available to Cereals, and to a lesser extent Pulses and oilseeds.

Table 4.4 shows for the eight simulations the estimated changes in capital dedicated to the industries in Group A and Group B.  The first thing to note about this table is that the elements all represent very small changes in capital stocks.  However, it should be remembered that the expansion of WA's agricultural processing sectors by $1 million caused only a relatively minor disturbance to the primary agricultural sectors (compared to their overall size), and so minor adjustments are to be expected.  The second thing to note is that within each group the adjustments to capital stocks sum to zero, demonstrating that within each group the capital stocks remain fixed. [6]

Table 4.4.         Adjustments to capital stock in agricultural industries (percentages)