In Defence Objectives and Funding (DOF), James Rolfe and I examined analyses that may be used in favour or against a small country such as New Zealand adopting variants of a balanced force approach. The analysis, while at a general level, indicated that the smaller is a country, the less likely it is to be optimal for that country to adopt a balanced force - i.e. each of an army, navy and (strike) air-force. However many other factors, such as co-operation with partner countries, ability and wish to contribute internationally to peace-keeping or other military campaigns and the general defence environment, impinge on this decision.

To place the proposed F16 purchase in context it is important (according to the analysis in DOF) to gauge what other defence outputs could be purchased with the sums projected to be spent on an air strike capability. This Part uses the Total Expense figures for output class D11 (Air Combat Forces) contained in the Report of the NZDF for the year ended 30 June 1999, and compares them with Expense figures and outputs for certain other output classes. This gives some indication of the extra services that could be purchased elsewhere through NZDF if a decision were taken to discontinue the air-strike capability.

However it should be noted that this analysis is still preliminary in that the only figures that have been used to derive these estimates are those contained in the Report of the NZDF. A more complete analysis of defence options would be required to assess the trade-offs between obtaining air-strike capability and strengthening force elements for a given level of expenditure. Also, such an exercise would be required to assess the benefits of switching (or retaining) expenditures across alternative force elements.

To simplify the analysis, we take the case where the air-strike capability (output class D11) is discontinued entirely. We also assume that other output classes exhibit constant returns to scale. This means that we assume that outputs within a certain output category will increase pro rata given the increase in total expenditure within that category. In practice, this probably understates the benefits that might be gained by redirecting expenditure in this manner, since each output class probably involves some fixed costs that could be allocated across higher expenditure in that class. This conservative approach appears warranted, however, given the major implications that might accrue from deciding to discontinue an output class such as D11.

Output Class D11

The total expenditure for output class D11 (Air Combat Forces) in the year to 30 June 1999 was $196 million. We use this figure as the expenditure available to increase expenditure in other output categories.

The $196 million figure comprises the following:

Expenditure Category	$ million
Personnel	54
Operating	53
Depreciation	30
Capital Charge	59
Total	196

Each output class has a different ratio of capital-related expenditures (depreciation plus capital charge) to total expenditures. In the case of D11 this ratio is 45%. In the figures for other output classes below, we concentrate principally on the aggregate expenditure figures, but note where the capital:total expenditure ratio differs significantly from that of output class D11.

D2 Naval Combat Forces

The capital:total expenditure ratio for D2, at 53%, is higher than for D11. Further, this total includes one frigate which has been depreciated quite fully. This suggests that an extra frigate would be more costly than indicated purely by the aggregate numbers (since capital-related charges would be higher for a new frigate than for the average of the existing frigates). I do not have the break-down of figures to accurately assess the effect of this factor. It indicates, however, that less than 1½ frigates would be able to be purchased for the expenditure currently assigned to D11.

D7-D9

Force Troops (D7)
Army Ready Reaction Forces (D8)
Army Training & Rotation Forces (D9)

Troop forces, with expenditure totalling $56 million, provide specialist (army) general support to operational forces. Components (and associated expenses) include:

• Strategic reconnaissance (SAS) force ($26 million)
• Strategic communications force ($12 million)
• Strategic movements force ($9 million)
• Force intelligence ($2 million)
• Force Military Police ($6 million)

Army ready reaction forces are for the conduct of land operations. The largest element (in terms of expenditure) within this output is combat forces (output 8.2) involving the provision of infantry and armoured units for the conduct of land combat operations with a force of up to one light infantry battalion group size (or brigade group size given longer preparation time). The total expense for output D8 is $274 million.

Army training and rotation force expenditure totals $128 million, providing reinforcement elements to sustain the ready reaction force.

Combined expenditure on these three inter-related areas is $458 million. The D11 expenditure is 43% of this amount. This implies, that if there were no additional forces required to supplement the extra personnel (noting that support forces are already included in output classes D8 and D9) the ready reaction force size and complementary units (D7 and D9) could be increased in size by approximately 40%.

In practice, however, fixed wing and rotary wing transport forces would also have to be increased. Together, expenditure on these two output classes (D13 and D14) totals $189 million, although it is doubtful if pro rata increases in these output classes would be required for any given increase in army personnel. However if a pro rata increase were required, the increase in outputs D7, D8, D9, D13, D14 for the foregone D11 expenditure would be 30%. Given the assumptions made, this is likely to be an under-estimate of the increased "army" and associated forces that could be obtained if an air-strike capability was discontinued.

An interesting contrast between output classes D7-D9 and output class D11 is that the capital:total expenditure ratio for D7-D9 is 28% (compared with 45% for D11). Thus the capital expenditure required to enhance the "army" as indicated above is less than that required to enhance the air combat force.

As a corollary, personnel expenditure in D11 is just 28% of total expenditure compared with a ratio of 49% for "army" expenditures (D7-D9). This extra personnel expenditure within the army may be seen by some to have public policy benefits compared with the higher expenditure on imported capital items within the air combat force.

D12 Maritime Patrol Forces

Current maritime patrol forces are based around two Orion aircraft being operational at all times, with a third being readily available for search and rescue operations. Three other Orions are on stand-by, maintenance, etc, with five crew currently available (increasing to six in 2001). Total cost for this output class is $88 million. This implies that this function could be more than doubled with foregone expenditure from D11 if desired.

The capital:total expenditure ratio for this output class is 39%. This is a little lower than for D11 but may again be affected by equipment of differing ages.

Commentary

The exercise in Part 3 of this report has to be accompanied by an analysis of the benefits of the various categories of expenditure before a final decision can be taken as to which type of expenditure(s) constitute "best value for money". The paper, DOF, suggests frameworks for undertaking such an analysis. Incorporated in such an analysis are not just "pure defence" objectives, but also "foreign affairs" objectives".

For instance, at one extreme if it were considered that the forces in these output classes were unlikely to be used in actual combat, foreign reactions to this country's choices may still impact on New Zealand's decisions. One set of countries may react favourably to New Zealand if the expenditure is used for an Air Combat Force (D11) while others may see "Army" expenditure as preferable. Depending on the manner in which New Zealand wishes to project itself externally, it may have to take such "atmospherics" into account.

It is likely, however, that at least some of these force elements will be committed within a realistic time-frame. Some information as to "value" of the various service elements can be gleaned from the implied existence of constraints on deployment over recent years. In this respect, army deployments - e.g. to prolonged deployment in Bougainville - have had to be curtailed, indicating that there is positive value in raising the ability of the ready reaction force to sustain its involvement externally.

A similar analysis of naval, air combat and maritime surveillance constraints should be made to make a judgement as to which the most pressing (and hence highest value) constraints are that should be relieved through redirection (if any) of defence expenditures.

ADDENDUM 1: Updated NPV Estimates

Since the robustness tests reported above were prepared, NZDF has revised the spreadsheet supplied to me to incorporate newer cost estimates of a number of categories. These have changed the baseline savings estimates of the A/B option over the C/D option in NPV terms as follows:

Original spreadsheet NPV Savings:	$351 million
Revised spreadsheet NPV Savings:	$342 million

The difference between the two figures is minuscule given the estimated size of savings, and given the much greater variability described in some of the robustness tests reported above. I have therefore not revised the figures in the robustness tests.