Are delays in tariff-reduction programs ever economically rational?

Employers and unions have differing opinions about the merits of a minimum wage rise on both employment levels and living standards

By Neville R Norman

Trends in world trade policy

Two striking features of trends in world trade policy since about 1990 are:

Significant, often spectacular, reductions in announced and actual nominal tariff rates; and
Delays and reversals in the actual tariff reductions and often the replacement of import-duty protection by other protective measures.

The first trend seems fully in accordance with economist textbook demonstrations that freer trade enhances national and world welfare. The tariff reductions arose from unilateral actions, multi-lateral agreements, such as China’s requirements to attain membership of the World Trade Organisation (WTO), and the succession of bilateral free-trade agreements (FTAs) between pairs of countries and trading blocs. The most spectacular recent example of significant reduction in tariff rates must be China’s near-removal of very high nominal tariffs. In 1996, import duties applied to foreign motor vehicles entering China were 80 per cent and 100 per cent, depending on engine capacity; while more recently they sit at 12 to 15 per cent. Australia’s car import duty rates were reduced from 57.5 per cent in the late 1980s to 10 per cent at present.

The second trend appears as a politically driven contradiction of the first trend, with no apparent foundation or support from economic analysis. A clear example of delay in the pre-announced program of tariff reductions occurred in the early life of the Howard Government in Australia. In 1998, substantial delays were applied to textile, clothing and footwear tariff reductions and to those applying to motor vehicles. Much earlier, the Whitlam Government’s 25 per cent across-the-board tariff reductions in July 1973 were followed by exchange rate, tariff and quota responses within two years that effectively reversed the tariff reductions.

The gaps in standard economic analysis

Standard economic analysis does not accommodate either time delays (or any specific chronology), or adjustment burdens imposed on specific sectors, regions, businesses or individuals. Accordingly, any significant emphasis upon those adversely affected by tariff-reduction programs and policy responses designed to address their plight can easily be branded ‘purely political’ responses that delay reaping the consumer benefits of lower tariffs. Press reports from journalists and some economists following the Howard delay decisions in 1998 said exactly this. Somewhat grudgingly, it was said that such delays might be the only political course to attain the goals of eventual free trade. We can call this the ‘adjustment burdens and adjustment costs’ political justification for delays in tariff reduction programs. In a broad sense it is also an economic argument, if one accepts that the political alternative is a return to the starting point before tariff reductions, as in the Whitlam example cited above.

Attempts to fill the gaps

Some attempts have been made by economists to incorporate this adjustment burden argument into formal economic models. In many ways, economists are responding to a huge challenge laid down by the prominent Canadian economist, Harry G. Johnson, whose writings and seminars dominated international economics in the 1960s and early 1970s. Johnson claimed that while politicians accepted the case for delaying tariff reductions, such policy responses were not derivable as policy-optimal from any formal economic models. He issued this challenge at international conferences, such as the 1969 Monash International Trade Conference organised by Ian MacDougall and Richard Snape, and the 1972 Cambridge World Development and Trade Conference organised by Oxford economist Paul Streeten. Johnson’s challenge does not seem to have appeared so bluntly anywhere in print.

One approach is to ascribe time paths to potential economic activity, with three relevant scenarios:

A control state where tariffs are left unchanged. Overall economic activity (say, real GDP), product prices and measure of welfare stay constant or grow at some pre-determined growth rate;
A radical one-off elimination of tariffs, with significant short-term adjustment costs and delays, as immobile and specialised resources are retrenched from import-competing sectors and take time to move to, or be retrained for production in, the efficient/export-competing sectors, if ever. Real national production reaches greater heights, eventually, than under the control scenario, the difference being the long-term gains from trade, which incorporates the main message from standard (static) economic analysis of tariffs in a more realistic dynamic setting; and
A staged or sequenced set of partial tariff reductions that takes longer to attain the static gains from trade. However, this strategy cushions the adjustment costs and makes it more ‘politically’ acceptable.

Some writers have explicitly allowed for adjustment costs. However, such factors are not part of the conventional or accepted literature. An important ingredient of any time-based approach to policy analysis is to incorporate present values by applying discount rates to time-specific economic variables. This procedure has the immediate consequence of favouring strategies that delay the bad news (adjustment costs) even though part of the good news (gains from trade) is also delayed. It also demonstrates how impatient policy propensities (high discount rates) will justify delayed rather than full-throttle tariff reductions. In formal terms, the delays are economically optimal, at least from the standpoint of individual nations. There are important pointers here for negotiators of free-trade agreements and also for the role and relevance of formal economic analysis.

There is another aspect of tariff reduction sequences that offers a completely different potential economic justification for delaying tariff reduction. It is not based on political responses or adjustment delays, or even significantly on present-value considerations. Rather, it arises from a close treatment of business investment responses to the alternative tariff-reduction scenarios. Once again, it is not easy or not possible to incorporate such aspects into standard trade and tariff theory. It is important to understand why this is the case.

Assumptions of classical tariff theory

Classical tariff theory implicitly supposes that the capital stock is fixed and given in what is ostensibly a stationary-state economic model driven by comparative static policy experiments. There is by definition zero business investment activity, other than capital stock replacement and maintenance. Consistently, there is no incentive for the perfectly-competitive firms to embrace profit-seeking extraneous activities as free entry negates any potential advantage they might seek. A broader world of multi-national firms operating in oligopoly markets with product differentiation and entry barriers offers richer opportunities. In this setting, there is potential for cost-reducing aggressive business investment and technology strategies to be adopted under the ‘delay’ scenarios to a fuller degree than under the full-throttle scenario. This is exactly what firms like General Motors argued to support the delays in tariff reduction in Australia in 1998. An important further consequence of the business setting we envisage is that costs and prices of (imperfectly-) competing products can be different, both from each other and over time. This feature is not consistent with standard tariff theories that invoke the ‘law of one price’ for all relevant products in every relevant market.

The case for modelling costs

Economists and other observers are wise to be circumspect about the arguments advanced by protected interests for delaying tariff reductions. This is why a return to modelling the full consequences of the tariff delays seems highly desirable. The two main questions that economists should ask are:

Would the cited business investment have happened anyway under the more radical reduction scenario?
Would the cited business investment detract from other/better uses of the same funds in other parts of the economy?

In the Australian case, companies like General Motors gave clear answers to the Australian Government. Firstly, they argued that the business investment would not otherwise have happened, an assertion supported by confidential board minutes and research papers. Secondly, if the investment had not gone ahead in Australia, it would not have released funds for any other investment in Australia. This second observation reflects one of many significant consequences of incorporating the activities of multi-national firms into the analysis. Even accepting both these representations, economists have further questions:
3. How extensive and sustained are the cost reductions flowing from the business investment that would not otherwise have occurred?
4. Are the welfare benefits to consumers arising from the cost reductions associated with the induced investment sufficient to offset the detriment to consumers flowing from the delays in the tariff-reduction program?
This is exactly the framework of the analysis set out in the paper to the National University of Singapore-University of Melbourne Symposium.

Clearly, there are some demanding tests to be passed before this investment-based argument turns out to be welfare increasing. In summary, there need to be positive answers to the four questions posed above, and significant quantitative cost reductions associated with the delay-induced investment, if the tariff delays are to be justified. That said, if all the tests are passed, we have established a new argument for potentially optimal delays in tariff reduction programs that has nothing directly to do with adjustment delays and burdens.

The model treats the policy-adjustment phase as one large period. There are three relevant avenues of final expenditure within the home country:

Outlays on the protected manufactured products, designated X;
Outlays on rival (product-differentiated) foreign-produced goods, denoted Y; and
Outlays on other products, called Z.

The tariffs applied to Y are either reduced in accordance with a pre-announced timetable, or are frozen for a time and then reduced later. We call these policy alternatives the ‘reduction’ and ‘freeze’ (or ‘delay’) scenarios.

The essential difference between these two scenarios is that we allow the level of production costs (and product prices in X) to become lower under the delay scenario, through the cost-reducing effects of the induced business investment. That carries consumer benefits associated with higher tariffs than would otherwise be the case. Anyone drilled in conventional tariff analysis would find this result (consumer benefits from higher tariffs than otherwise) difficult to accept, but it has already been established in the literature by economists Helpman and Krugman. The result is also consistent with recent econometric work directly investigating price responses by import-competing firms in the case of cost and global pricing influences. The criterion equation developed is whether this induced investment benefit does or does not compensate for the alternative-policy loss of welfare arising from delaying the benefits to consumers form the earlier reduction in tariffs under the reduction strategy. This criterion can be transformed to ask a more direct question: What degree of investment-induced cost reduction is required to justify delaying the program of tariff reduction? The model provides the framework for these questions to be answered and for trial or actual numbers to be inserted to gauge quantitative answers.

Conclusion

The general conclusion is that, subject to a sufficient induced cost-reduction, the case remains open for delays in pre-announced tariff reduction programs to be policy optimal. Moreover, this case does not depend on adjustment delays and burdens, which supported previous demonstrations that tariff delays may be policy optimal. In many ways, we have developed a more strictly economic, rather than political or reactions-based, set of circumstances that may justify the delay.

The message here is not that delays in tariff reduction programs are justified. It is that a set of questions can be developed that open up this possibility. The questions are specific and demanding. Unless the welfare gains from cost reductions induced by the business investment are sufficient, and unless the business investment would not have arisen without the incentive effects of the tariff delay, the case is not made out. Indeed, the way in which classical tariff theory is set up, these questions cannot even be asked, let alone be answered.

Top ^

A shortened version of a paper given at the Conference on Regulation, Governance and Markets in Small, Open Economies organised by the Department of Economics, University of Melbourne and the National University of Singapore on 14-15 December 2006. The full paper is available here.

Associate Professor Norman is a member of the Economics Department of the University of Melbourne.

Volume 2 OCT 2007