LP Modelling of NZ Dairy Production – A Counter Intuitive Outcome?

Is a change of direction of NZ dairy possible?

Based on current practices, the short answer is no.

The issue is that for the past decade, an increasing number of dairy farmers have moved to much more intensive farming systems and/or farmed in marginal areas to chase production volumes in the attempt to increase revenues.

However, this has not been without significant concerns, as intensive production and extensive growth have resulted in escalating environmental concerns.  Put simply, environmental constraints are likely to mean that dairy will be unable to ‘grow its way out of debt and back into profitability ’.

OK, so what is a solution then?

While intensive and extensive dairy has been criticised from an environmental perspective, it is not well understood that it raises a serious questions from an economic perspective as well.

The secret of New Zealand’s traditional comparative advantage in dairy has been grass, which is a low cost feed source.  By chasing production there is the very real likelihood that a significant minority of farmers are pushing the production boundaries beyond where it is economically sensible to do so.  In such a situation the sensible thing to do is to de-intensify as marginal costs will fall faster than marginal revenues, thereby helping to restore profitability. The $64,000 question is whether this will happen over enough farms quickly enough to restore sector profitability and thereby avoid ongoing falling land prices.

How can the GSL model help matters?

Unlike other farm system models, the GSL model is based on a linear programming format rather than a benchmarking formation, which allows the GSL model to optimise resources allocated.  In particular, it can show a farmer where the profit maximisation point is by matching marginal costs and revenues.

Put simply, this means that farmers can be assured that the last cows they have are making them money rather than costing them money.

Who has used the model and what were the results?

MAF has commissioned GSL to model the impact of emission charges on farm profitability; and to test whether New Zealand dairy farmers are producing before, at, or beyond the ‘tipping point’ whereby applying additional inputs (e.g. cows, imported feed, etc) decrease farm profitability.

Initial modelling was conducted in 2007 and used MAF farm monitoring data. Subsequent modelling of three actual Waikato farms was undertaken in 2010.  Both studies showed the ability to substantially reduce herd numbers whilst simultaneously maintaining or even increasing profitability.

The second study produced the starkest results, which indicated the ability to achieve:

  1.    Large and positive improvements in farm income (up to +53%),
  2.    significant reductions in costs (up to -30% ),
  3.    useful decreases in carbon emissions (up to -7%), and
  4.    Increases in production (up to 6.5%).

Was it just the case that these farms were especially poor performers?

While on farm performance of these farms was generally poor, it was not atypical (e.g. kgMS per cow ranged from 241 – 313 [inc. use of supplements] and DM production ranged from 9,700 – 13,700 kg ha).

DairyNZ data states that New Zealand’s average per cow is currently around 350 kgMS.  Given this average includes higher intensity farms (that with supplements, produce well in excess of 400 kgMS per cow) it also implies that many farms have per cow production of less than 300kgMS; which means the per cow production figures from three Waikato farms modelled above were by no means outliers.

Indeed, the Waikato average production per cow in 2009 was 307 Kg from 2.93 cows per hectare.

So is the problem not enough grass or too many cows?

The problem is arguably both.

At the risk of oversimplification, the modelling work suggests two broad

categorisations of ’problem farms’, namely:

  1.    Low efficiency farming:  farmers fail to grow sufficient grass, so attempt to maintain production by using supplementary feeding regimes. These farmers would benefit from growing more grass to increase per cow production towards the 400 kgMS possibility frontier and thereby eliminating the need for supplementary feeding.
  2.    High intensity farmers: farmers that feed substantial volumes of supplementary feed to maintain high stocking rates and high rates of per cow production. Modelling results suggest that these farms tend to be producing at a point where

marginal costs are greater than marginal revenues, which implies de-intensifying will lead to the counter-intuitive outcome of improved profitability.

How does the on farm cost structure work then?

Analysis of the farm level results suggests that the operating costs can be split into per cow costs and dry matter costs.  Costs per cow show substantial variance at low levels of production but smooth to being relatively flat with increasing production.  Dry matter costs are very different and can be split into three regions based on production per hectare:

  1.    < 450 kgMS per HA – dry matter costs are essentially free i.e. feed comes from what the farm produces naturally given its soil type and climate. This low cost grass based production system is what made New Zealand dairy production the most efficient in the world.
  2.    450-800 kgMS per HA – increasing production requires additional dry matter
  3. – typically from purchasing grazing and using fertilizer to increase pasture production.  For these resources the dairy farmer competes with other uses of these resources and with other dairy farmers.  Over this range costs rise with higher production intensity because cheaper sources of dry matter are used first and then more expensive forms. This is still, however, predominately grass based milk production.

  4.    >800 KgMS per HA – additional milk production comes from the purchase of supplementary feed.  The use of supplementary feeds has the following impacts:
    • Provided profit maximising uses of resources are already being made, the marginal cost of production, in almost all situations, exceeds the marginal revenue received.  If, however, resources are being poorly used supplements can mitigate the damage already done and provide an apparent, but false, positive return.
    • Additional capital is required for plant, equipment and more complex systems.
    • Feeding supplements changes cow behaviour that decreases pasture utilisation.
    • Farm management of pasture typically also declines.

If this model is so good how come farmers aren’t beating a path to your door?

At its most basic level the model is highly counter-intuitive and typically sounds ‘too good to be true’.  Critically, concepts such as marginal analysis and optimisation are not well understood (if at all) and tend to cut across existing orthodoxy or industry wisdom.