3. Implications for Economic Policy
3.1 Assessing the Need for Policy Change
The New Zealand economy, like any modern economy has in place a complex web of government interventions – taxes, publicly provided goods, regulations governing social and economic behaviour, enforcement mechanisms, etc.
Much of the complex web is aimed at facilitating income growth. In addition, governments intervene for a variety of other economic and non-economic reasons. The New Zealand government for example has a heavy hand in health, education and welfare. And it has to finance these and other interventions through the tax system. The design of these interventions also has the potential to impact on growth.
Since economic growth is driven by entrepreneurs innovating, the touch-stone for policy should be identifying and remedying situations where socially valuable innovation is more difficult than it needs to be. In other words, the objective of intervention should be to facilitate innovation by enhancing the ability, connectedness or incentives of entrepreneurs to spot and pursue socially worthwhile outcomes and to discourage them from pursuing socially undesirable outcomes.
There are a number of ways that we can identify areas where changes in policy can potentially improve the functioning of markets in contributing to income and income growth. Two key ways are:
- Empirical studies identifying what capabilities are conducive to growth.
- Analysing from a more theoretical perspective how markets and the institutions underpinning them may fail to deliver good outcomes.
These two ways are not without their difficulties. Often, these difficulties can by reduced by using the two methods together. In particular, empirical work will inevitably not give us a full picture. We can use empirical work to provide some "stakes in the ground" in terms of creating a map of what is important for productivity growth, and then use theoretical reasoning to bridge the gap between the "stakes" to give a more comprehensive picture19.
3.2 Empirical studies
Empirical studies can be used to identify at a broad level the capabilities that are associated with high income levels and growth rates. While these studies are not without their difficulties, they have the virtue of directly assessing the impact of various interventions on material well-being.
These studies can be used to assess what capabilities are important for economic growth. The capabilities that this statistical evidence suggests are important are summarised under heading 2.11 above.
The main disadvantage of these studies is that they cannot give evidence on all important capabilities, including in particular on those capabilities that are only observable at a low level of disaggregation (such as the need to have a register of real estate, as discussed in the real estate example above). This is an important omission, since (as the discussion above makes clear) many of the important capabilities will only be observable at this highly disaggregated level.
These empirical studies suggest that the New Zealand economy has institutional settings that are relatively conducive to promoting good economic growth (OECD 2003). It has a relatively stable and cohesive society. It has political structures that both allow institutions to be adapted and that are accountable to the public. It has an independent judiciary and a police force that largely acts within the rule of law. Corruption is low. The rule of law and property rights are well established, meaning that entrepreneurs can expect to be rewarded for successful innovation. Competition and consumer protection law tend to limit competition along undesirable margins and so foster competition along desirable margins. Markets are flexible, allowing entry of new businesses and exit of poorly performing businesses (McMillan 2004). And so on.
As a result, New Zealand markets are relatively effective at giving people the incentives, ability and capacity to pursue their own welfare and in doing so undertake socially worthwhile activities.
However, as the analysis above makes clear, just because markets are effective at any point in time does not mean that the institutions underpinning them do not need adjustment over time as the environment changes and knowledge improves. In addition, although the institutions are good, it does not mean they are as good as they can be. Studies comparing New Zealand with other countries suggest that many other OECD countries do better than New Zealand on some important dimensions (see for example Ministry of Economic Developmentet al 2007, especially page 7).
Empirical studies can also be used to assess whether interventions undertaken for other reasons are having an adverse and unnecessary impact on incomes. The design of these interventions has the potential to impact for good or ill on the operation of markets including entrepreneurial activity within them. For example, the method of remunerating tertiary education staff leads to a bias in staff quality away from market-relevant disciplines (Boyle 2007). Similarly, the New Zealand tax system has a relatively heavy bias towards capital income taxes (Ministry of Economic Development et al 2007, page 85). While these biases may be justified on other grounds, they are likely to be less than helpful for growth.
3.3 Theoretical Approaches
The second way of assessing the effectiveness of markets and the need for policy change to improve economic growth is to adopt a theoretical approach. For example, using neo-classical reasoning about market failures can be helpful, provided it is done judiciously. In particular, four cautions are warranted:
First, not all neo-classical market failures are bad for innovation. Market imperfections are necessary to foster innovation (Lipsey et al 2005, page 45-46). For example, economic rents20 are necessary to encourage innovation, and innovation may be fostered by head-to-head competition between oligopolistic rivals (Aghion, Bloom, Blundell, Griffith and Howitt 2005). We therefore need to make an explicit judgement about whether a particular market failure is likely to foster or inhibit innovation.
Second, neo-classical economics tends to abstract from many important capabilities, and therefore potential deficiencies in those capabilities. We need to explicitly address those capabilities in using neoclassical reasoning to identify policies to enhance economic performance. In particular it tends to underplay the difficulty of acquiring information that already exists and of interconnections between agents.21 For example, social capital is important in establishing trust and so facilitating low cost contracting.
Third, we still need to apply judgement about what is "good" innovation (e.g., we probably don't want the market for stolen goods to work better, and we probably don't want businesses to innovate by finding more sophisticated ways of misleading consumers).
Finally, the analysis will not typically identify how important the failure that is discovered is to growth.
Nevertheless, it is likely to be a useful basis for assessing whether the functioning of markets and government interventions can be improved, especially where empirical evidence is not available.
The (interrelated) characteristics of markets that may lead to poor outcomes – market failures22 - include:
- Uncertainty
- Incentives
- Spillovers
- Path Dependence
- Information Asymmetries/ Missing Markets
- Coordination Failures
- Non-rationality
These are discussed in more detail over the next few pages.
3.3.1 Uncertainty
The earlier discussion highlighted the fundamental problem that entrepreneurs face in dealing with uncertainty. While much of this is unavoidable, reducing the extent of uncertainty that entrepreneurs face allows them to focus their attention on dealing with the unavoidable uncertainties at the heart of innovation. Certainty, clarity and trust (important for ease of contracting) can be improved by establishing:
- efficient property rights and speedy and impartial enforcement of contracts.
- a stable macroeconomic environment to ensure that the most important and pervasive prices in the economy reflect fundamental value, and not, for example, short term business cycle fluctuations. Most OECD countries, including New Zealand, generally perform well on this score. However, for New Zealand, exchange rate volatility is a significant issue.
- an intuitive23, predictable and relatively stable regulatory environment.
- a predictable approach to changes in regulation.
3.3.2 Incentives
Regulation and other institutions are essential to ensure that it is easy and profitable for entrepreneurs to "compete at those margins, and those margins alone, that induce growing productivity", and hard to compete at the myriad remaining margins. Entrepreneurs' guiding light is profit, and they will compete and innovate along whatever margins are profitable, whether those margins are socially beneficial or socially destructive. "[If] the incentives reward piracy, then that will be the outcome" (North 2005, page 77, 84).
There are a wide range of policies aimed at ensuring innovation takes place along socially productive margins. For example, law, regulation and other institutions can help people and businesses overcome information problems and help prevent the creation of information asymmetries for welfare-reducing commercial gain. Examples include laws on corporate governance and consumer protection (Gans 2005, North 2005 page 76).
Beyond ensuring that only socially beneficial innovation is profitable, regulation can help ensure that entrepreneurs and businesses are actively encouraged to innovate along those margins that are socially beneficial. In particular, strong competition law, including law that promotes head-to-head competition between oligopolistic rivals, promotes innovation (Aghion et al 2005, Baumol 2002).
3.3.3 Internalising Spillovers
While a business profits from its contribution to successful innovation, it typically does not reap the full benefits of its discovery. As is discussed above, a new innovation creates new knowledge and the other capabilities needed to exploit it, which provides: (i) a higher platform for further refinement and diffusion; (ii) a platform for further discoveries; and (iii) a stepping stone to new, higher-value products and processes. The future result of this process in many cases will be unimaginable when the innovation is introduced. This applies to the many forms of knowledge generation, including innovation in production, processes, organisational form, marketing and distribution.
For example, as is discussed above, the discovery of electricity in the 18th century has allowed a large number of further inventions that have literally transformed the way that our societies function. Similarly, the development of the production line, often attributed to Henry Ford24, eventually transformed the process of manufacturing.
Likewise, when a business starts successfully producing a new product or exporting to a new market, it shows that the country has the capabilities and cost structure necessary to compete in those markets. This generates information that is valuable both to competitors and to other businesses producing similar products (Hausmann and Rodrik 2002). For example, in New Zealand, the people who first discovered that wine could be profitably exported provided valuable commercial information to all others who later set up their own vineyards and wineries.
In implementing innovations like these, each business also provides a training ground for its employees to learn the capabilities necessary to implement the innovation, which it may not capture if the employees leave to set up or join a competing business.
Thus, innovation typically results in new products, processes and capabilities that provide substantial benefits to society over and above the profits to the business – i.e., substantial spillovers. This provides a potential justification for government support for such innovation.
3.3.4 Path Dependence
As is noted in section 2.8 above, moving towards some part of product space rather than others is more valuable in terms of the number of new high value (in particular, nearby high income-content export) production opportunities that it opens up. So some development paths are more propitious than others in terms of the opportunities that they open up for future growth.
Sometimes, it will be commercially beneficial for businesses to pursue the most propitious path, but this will not always be the case. This is because each entrepreneur is interested only in the opportunities they generate for themselves, whereas society and the government are interested in the opportunities available to the economy as a whole and so all New Zealand entrepreneurs in the future.
There is therefore a potential case for the government to influence the economy towards parts of product space with better future growth potential, by encouraging some direction of development over others25.
3.3.5 Information Asymmetries/ Missing Markets
As Stiglitz 2002 (page 478) has pointed out, information asymmetries mean that "market failures are pervasive". Information asymmetries also mean that markets cannot always be relied on to produce good outcomes on their own.
Because innovation is about the creation of new knowledge, a particular form of information, information asymmetries are particularly important for innovation. There are likely to be missing markets, and the institutions that evolve to substitute for the missing markets may not improve matters. For example, information asymmetries can limit the development of capital markets. Financial markets to fund human capital development and start-ups, where collateral is limited, are likely to be weak (Stiglitz 2005, pages 25-27).
3.3.6 Coordination Failures
As is clear from the above, innovation involves the coordination of many different capabilities. While it will often be viable to undertake that coordination as a commercial venture, there will be other instances where it will not. For example, there may be such free rider problems that it is too difficult to coordinate activity, or once activity is coordinated, too difficult to extract sufficient commercial gain. In this case, government intervention can help26.
3.3.7 Non-rationality
The assumption that individuals will act in their own best interests is only approximately true. There is clear evidence from the field of behavioural economics that there are systematic biases and "imperfections" in individual behaviour (NEF 2005). There is also evidence that these biases are exploited by businesses (Cialdini 2001).
Even if this were not true, the complexity of the world means that individuals build mental models of the world and they behave as if that model were reality. Over time, societies build cultures which form a background to and influence individual decisions. Both individual beliefs and cultures may not reflect reality, meaning that innovation may not be constrained by reality, but by people's perceptions of it (North 2005, for example page 117).
3.4 Government Failures
As is noted above, ongoing active government intervention is inevitable in any modern economy, and certainly in an economy like New Zealand. Sections 3.1 to 3.3 suggest areas where changes to government interventions might yield benefits to innovation.
3.4.1 Rent Seeking
An important corollary of the need for ongoing government intervention is that a government strong enough to impose rule changes is essential. The paradox is that such a government is also strong enough to prey on the market for the ends of those who control the government.
This raises two potential concerns. The first concern is that the polity or the bureaucracy may undertake interventions that pursue their own interests. The second (and related) concern is that business may see the government intervention as an opportunity to divert some of the government's tax revenue or other resources to enhancing its bottom line (so-called rent seeking27). If scarce entrepreneurial efforts are diverted towards innovation aimed at extracting rents, they are diverted away from socially productive innovation. Thus intervention that raises the risk of rent seeking can be damaging to economic growth. This is a major concern, and it is essential that policy be designed to limit the incentive for and potential scope of rent seeking. It is better not to implement a policy change than to implement one that chances encouraging a culture of rent seeking.
"The government is hardly a disinterested party. The structure of political markets will determine whose voices are "heard" in shaping additional rules governing each market." Thus "…well functioning markets require government, but not just any government will do. There must be institutions that limit the government from preying on the market. Solving the development problem therefore requires the crafting of political institutions that provide the necessary underpinnings of public goods essential for a well-functioning economy and at the same time limit the discretion and authority of government and of the individual actors within government" (North 2005, pg 124, 85).
3.4.2 Poor Targeting
Improving an intervention requires us to be able to identify it and its deficiencies with sufficient precision that it can be adequately addressed. This raises a key issue. Government policy makers have the same bounded rationality and face the same radically uncertain environment as entrepreneurs. In fact their difficulties are greater, since the government is further distant from market sources of information. As a result they can never be certain that they have correctly identified an area that needs policy attention or have correctly designed and targeted an intervention to address the problem. Likewise, those implementing the intervention face similar difficulties. Furthermore, there is no market test to sift out poorly designed or implemented interventions.
This is not an argument for not intervening, since there is no a priori reason to suppose that the existing set of interventions is in some way superior to any new or revised intervention. However, it is an argument for:
- Designing the policy so that it is robust to poor targeting or poor implementation, and that it allows adaptation to changing circumstances over time
- Ensuring that those designing the policy have sufficient competence and knowledge of the relevant part of the economy to understand the rationale for intervention and to design the policy to be appropriate to the circumstances.
- Ensuring that the policy is sufficiently simple to implement that the officials charged with implementing it have the capabilities to implement it appropriately (this may imply simple implementation or officials with a sophisticated understanding of the policy).
- Subjecting each intervention to robust formal evaluation28.
There are questions about whether these conditions are always, or even normally, fulfilled with respect to industry policy in New Zealand.
3.5 Costs and Benefits of Different Interventions
Where and how to intervene must be decided on the facts of the specific situation. This requires an assessment of the costs and benefits from doing so. The costs include the resource costs of intervening, the costs the private sector faces in adjusting to the intervention and the costs of any rent seeking induced (directly or indirectly) by the intervention. The benefits arise from promoting innovation by better providing people with the ability, connectedness or incentives to spot and pursue socially worthwhile outcomes, as discussed in sections 3.1 to 3.3.
It is important that, wherever it can, the government adopts a transparently and soundly reasoned and evidence-based approach to intervening, for two reasons. First, it increases the chances that the policy is appropriately targeted. Second, provided the government takes – and develops a reputation for taking – such a reasoned and evidenced based approach, policies can be relatively robust to rent-seeking. Under these circumstances, lobbying efforts are likely to be concentrated on adducing evidence about the impact of potential policy changes, an approach that is supportive of good policy.
3.5.1 Broad-Based Policies
For broad-based policies (policies have an impact across large parts of the economy)29, we can often generate an evidence base using international statistical evidence such as cross country studies (for example, OECD 2004) to provide a foundation for appropriate policy. This allows us both to assess the nature of the appropriate intervention and its probable impact. In addition, we can normally design the policy so that its ongoing implementation (after the initial design and implementation) can be safely delegated to people who do not need to understand the reasons for the policy. (For example, those assessing whether some expenditure qualifies for a R&D tax credit need to know only the relevant legislation, not why the legislation was implemented).
Because broad based policies have an impact across large parts of the economy, even small changes to them offers the potential for substantial gains (and losses). New Zealand has broad-based policies that approach international best practice (OECD 2003). Nevertheless, there are still opportunities for further measures along these lines. For example, the recent introduction of the R&D tax credit, should, on the evidence available, lead to further gains in economic growth (Davis 2006). The Economic Development Indicators Report 2007 (Ministry of Economic Developmentet al 2007, especially the summary on page 7) highlights a range of interconnected areas where New Zealand's economic performance is mediocre by OECD standards. While these do not necessarily indicate that further intervention is required, they are deserving of further consideration. For example, the Report points to weaknesses in New Zealand's capital markets30.
Thus the potential net gains from further evidence-based improvement of broad-based policies are substantial and they should continue to be pursued.
3.5.2 Fine-Grained Policies
The description of innovation and growth in the first half of this paper makes clear that many growth opportunities will actually occur at a lower level of disaggregation than can normally be assessed by statistical data. Many of the capabilities that might be needed to take advantage of these potential growth opportunities can similarly only be assessed at this lower level of disaggregation. This can be illustrated by the example of the housing market discussed earlier. Some of the capabilities cited are provided by the private sector (e.g., real estate agents) and some are provided by the government (for example, the registration of title and of charges against the title).
Since business will have the incentive to provide any capabilities that are required, the binding constraint to innovation is likely to be capability requirements that require government action. Fine grained policies31 to address them, if they can be implemented effectively, are therefore potentially important in promoting growth. This suggests that the government should seek to find a practical way of intervening at this level.
Typically, independent empirical information is not available about such opportunities. In order for the government to identify those products and whether it is socially beneficial to provide them, it must gain a detailed knowledge of the sector, knowledge that can likely only be obtained directly from the sector involved32. This requires a process of discovery, a process that necessarily involves engagement with the private sector actors involved. One way of engaging is to choose a unit for engagement that is likely to have some commonality in the capabilities it needs. This might be at the level of regions, sub-sectors, products, or even individual businesses (referred to from here on as "sectors").
Hausmann and Rodrik express this need for engagement this way. "The government has only a vague idea at the outset about whether a set of activities is deserving of support or not, what instruments to use, and what kind of private sector behaviour to condition those instruments on…. An industrial policy that is cognisant of the government's lack of omniscience has to be constructed as a system of discovery about all those sources of uncertainty. It requires mechanisms for eliciting information about the constraints markets faced, and hence close collaboration between the government and the private sector" (Rodrik 2007 page 38, Hausmann and Rodrik 2006).
This sectoral engagement opens up the additional possibility of coordinated action between the government and business, which (rent seeking aside) is likely to yield greater benefits than could be generated if both worked independently. Rodrik states that a strategic partnership between the government and industry will facilitate a process of self-discovery, with the aim of "learning where the binding constraints lie and on eliciting information on the private sector's willingness to invest subject to the removal of those obstacles". The objective is to engage in a genuine cooperative agreement where each party is undertaking the sorts of activities it normally undertakes, but in a way that shares information and coordinates actions (Rodrik 2007, page 40).
The potential benefits include the following:
- the government can adjust its funding of industry good research to complement commercial research being undertaken by businesses in the sector. In doing so it can encourage businesses both to do more research and to coordinate their research so that the benefits can be spread more widely.
- the government can adjust its funding of higher education (as the Irish and Finnish governments have done) so that students are encouraged to study the skills that are needed to support the direction of the country's growth strategy or that businesses identify they need to expand.
- the government can alter regulation that is inappropriately interfering with the ability of a business to prosecute its strategy effectively.
In practice, the government already makes decisions that have fine grained effects - but often with the various decisions fragmented across parts of government and taken in an uncoordinated manner. Or it simply reacts to decisions of the private sector.
For example, decisions by ports regarding expansion and mergers have obvious flow-on implications for the road and rail infrastructure needs of the surrounding area. But the decisions by ports are generally made in isolation from those broader implications. And the government needs to either try to predict ports' decisions or react to them ex-post. Neither approach seems ideal.
Similarly, parts of the government are making choices regarding tertiary education funding and research funding. They can be coordinated and informed by a broader development strategy, or devolved to operational agencies to make on the basis of their sector knowledge – with varying degrees of sector input.
Beyond this, changes to other capabilities, such as regulation, institutions, networks etc, may also be required to facilitate economic growth. Lipseyet al2005 (Part IV, especially page 517ff) present a fuller exposition of these issues, and in particular identify two such approaches.
First, changes in technology will often require changes in what Lipsey et al term the facilitating structure to be effective. (The facilitating structure is defined by Carlaw in Procter et al 2008. Broadly it is what we identify as capabilities above). First, changes in technology may necessitate changes in government regulation to support them. For example, New Zealand has recently altered the way it allocates radio spectrum to accommodate changes bought about by the ICT revolution, notably digital television. Secondly, public policy can assist in changes to the facilitating structure that are provided by the government, perhaps because they are subject to major externalities. For example, it may involve changing education to provide training in using the new technology. It may also involve changing elements of publicly owned infrastructure to make best use of the new technologies.
As well as reacting to changing technology, the government may also act to create changes in the facilitating structure that will foster innovation. Examples of such policies include attempts to integrate some university, government, and private-sector research activities; attempts to create technology information networks; and attempts to change private-sector attitudes toward adopting new or different technologies. Furthermore a government can attach structural conditions to funds given to businesses to assist them to develop technologies that they would have developed anyway – for example research labs – in a slightly different way.
Thus there is potentially a good case that fine-grained policies can yield gains over and above those that are available from broad-based policies.
However, such engagement on fine-grained polices poses particular challenges First and most important, they exacerbate rent seeking concerns, for a number of reasons:
- The Benefits of Rent Seeking to the Business. With fine grained policies, the benefits of rent seeking tend to be relatively large for the businesses involved, because it can capture most of the benefits, and because capturing the benefits gives it a competitive advantage over its rivals. This is not true to the same extent for broad-based policies.
- The Fiscal Cost to the Government is relatively small. Only a few private sector players are involved, so the assistance provided is not transparent and there is limited impact on the Crown's balance sheet.
- Political Processes, unlike market processes, can often subsidise losers at the expense of winners.
These are potentially serious problems, which apply to all policy, but particularly to more fine-grained policies. Because of their potential for undermining productive innovation we should be cautious about pursuing fine-grained policies even where these effects can be mitigated.
Second, identifying the policy requirement and the appropriate policy response, and assessing the resulting costs and benefits will be specific to the particular circumstances of the issue being addressed. This involves a range of skills, including in particular a good understanding of the issues discussed above and good judgement to apply them in a robust way. This clearly will require highly competent officials, which will in itself limit the scope to implement such policies. It also requires deliberate ex-post evaluation to ensure the hoped for gains are in fact being realised. Attention should therefore be focussed on a limited number of areas where the net gains are likely to be greatest. Similar considerations apply to implementation of the policy33.
Third, and related to the previous point, the resources necessary to implement policy change, both politically and technically, are limited. It is therefore important to focus on policies with a big potential payoff. In particular, in undertaking fine grained policies, it is important to assess the opportunity cost in terms of foregone implementation of other policies, particularly broad brushed ones. An excessive focus on fine grained policies risks the appearance of action rather than the reality.
3.6 Designing Policy to Maximise Net Gains and Limit Risks
The greater the potential costs from rent seeking, the less the potential net benefit of any intervention, and so the more conservative a government should be about implementing such policies. The design of policies and the decision to pursue them needs to take account both of the potential benefits from engagement (discovery) and of the potential costs, including in particular their potential to divert entrepreneurial effort to rent seeking.
The policy challenge therefore is to focus on areas where the potential gains are greatest and to design policy and institutions to limit these costs.
In terms of institutional and policy design, the objective should be to allow market signals to drive policy implementation while limiting the rewards to, and increasing the costs of, rent seeking. In particular this implies the following general principles:
- policy should focus on areas where there is substantial potential to contribute to income levels and growth.
- the focus should be "broad-based" with respect to involvement of particular businesses. Any contracts with individual businesses should reward external benefits, not the business itself.
- commercial risk should be left in the hands of the private sector.
- the particular intervention chosen should be designed to be robust to a number of possible potentially valuable outcomes, in recognition of the radical uncertainty facing any policy intervention34. At the same time, it should protect against socially undesirable outcomes. For example, this points to the desirability of principles-based, rather than rules-based, regulation (while also recognising that the latter may require greater technical competence in implementation).
- policy principles should be set by the polity, but implementation of those principles should be delegated to non-elected officials who apply the principles to particular cases independently. The reason is that rent seeking concerns are sharpest at the implementation stage. Non-elected officials are typically both less responsive to rent-seeking by individual businesses and more easily disciplined if they stray from the approved principles. This approach is now well established in New Zealand practice – for example, with tax law, state financed commercial activities (State Owned Enterprises) and monetary policy.
- policy should be designed by officials with a good understanding of the rationale for the policy, as outlined in this paper, and a good understanding of the parts of the economy to which the policy will apply.
- policy should be designed so that as far as practicable, its ongoing implementation does not require an understanding of the policy rationale. Where this is not feasible (as is true of much of industry policy) the officials tasked with implementing the policy need to have competence and judgement to understand and implement the policy appropriately.
- the officials' decisions and decision processes should be subject to public scrutiny and officials should be accountable for applying the principles in a principled and non-corrupt way.
- institutions and processes should be put in place to reinforce this accountability – for example, watch-dogs (analogous to the Independent Police Complaints Authority and the Privacy Commissioner)
- institutions should be put in place to reinforce an evidence based approach and to counter the tendency for governments to continue backing losers - for example, mandated independent evaluation and sunset clauses.
Back to Top