National Economic Review
National Institute of Economic and Industry Research
No. 67 November 2012
The National Economic Review is published four times each year under the auspices of the Institute’s Academic Board.
The Review contains articles on economic and social issues relevant to Australia. While the Institute endeavours to provide reliable forecasts and believes material published in the Review is accurate it will not be liable for any claim by any party acting on such information.
Editor: Kylie Moreland
© National Institute of Economic and Industry
This journal is subject to copyright. Apart from such purposes as study, research, criticism or review as provided by the Copyright Act no part may be reproduced without the consent in writing of the relevant Institute.
The economic impact of public events
Ian Manning, Deputy Executive Director, NIEIR
The present paper examines a small and ill-defined area of government–private sector interaction: the organisation of public events. In particular, events that receive subventions from the public purse and that draw patrons from outside as well as within the subsidising jurisdiction are considered. Neoclassical economics asserts that economic policy should concentrate on leaving decisions to markets. As applied to the analysis of events, the neoclassical policy recommendation is for reliance on user charges except where there are non-economic reasons for free or subsidised provision. However, NIEIR’s analysis of event provision shows that this recommendation relies on assumptions that have not applied in most parts of Australia over the past 30 years. Instead, subventions to events that increase tourist visitation can increase incomes and employment. Thus, the case for subventions becomes one of priority against other expenditures, preferably exercised in the context of a coherent strategy for the future.
Mixed economies and government– private sector interaction
When the United States went into ideological overdrive following the end of the Cold War, it let loose a wave of propaganda for market economics. Soviet central planning and government ownership of business enterprise were discredited. Economies should, therefore, be reformed to the opposite extreme, with government curtailed to providing law and order and all else left to the private sector. Some of the most effective propaganda involved the promotion of sophisticated neoclassical economic modelling, which purported to show that a completely privatised economy delivered the best of all worlds. In countries dominated by the American way of thinking, this modelling became the standard by which economic policies were assessed. Its intended target was not the Soviet-style centrally-planned state but the mixed economies inherited largely from the wave of socialist reforms that followed the Second World War.
In Australia’s case, it was conveniently forgotten that the economy had been mixed ever since the 19th-century colonies found that private enterprise was unable to provide them with the infrastructure they required. For the most part, academics and treasuries influenced by the neoclassical wave of thought deliberately assumed away the insights of Keynes’s generation, which supported mixed economies. One of these was the finding that markets are generally ‘imperfect’, which is a polite way of saying that they cannot be relied upon to generate an optimum allocation of resources as described by neoclassical theory. This is not to claim that government action can be relied on to improve matters: far from it. However, there are occasions when governments have opportunities to increase employment and incomes. It is arguable that they should be on the lookout for such opportunities and take advantage of them when they arise.
Since its foundation in 1984, NIEIR has remembered the findings of the economists of Keynes’s generation and, accordingly, has stood apart from the neoclassical economic models favoured by the American economic evangelists. The intellectual problem with the neoclassical models is their high level of abstraction and the plethora of assumptions, many of which contribute directly to their pro-market findings. It should not be surprising that when attempts are made to apply them to policy problems, highly abstract models can generate misleading results. It is inevitable that economic models should simplify the intricacies of the mixed economy in which governments, privately-owned business and non-profit organisations have basically complementary tasks but can sometimes be in (necessarily imperfect) competition, but important that the simplifications capture the major causes and effects rather than being assumed from ideological principle.
The commitment to the vision of a purely private market economy has diverted attention from the realities of government–private sector interaction. These differ industry by industry. Although all industries rely on government for the enforcement of contracts and property rights, the dependence of the mining sector on the administration of exploration and mining licences differs from the dependence of the finance sector on the administration of debtors’ and creditors’ rights and its ultimate dependence on government as lender of last resort. Again, there is general reliance on governments for the provision of an educated workforce but skill requirements are frequently industry-specific. Despite the targeting of government-provided physical infrastructure in privatisation campaigns, one major industry, road transport, is, to this day, completely dependent on infrastructure owned and managed by the three levels of government. Others depend on infrastructure that is to various degrees government provided, guaranteed or assisted. The mixed economy refuses to go away.
Event organisation as an activity
This article looks at a small and ill-defined area of government–private sector interaction: the organisation of public events. In particular, the paper considers events that receive subventions from the public purse and that draw patrons from outside as well as within the subsidising jurisdiction. Public events in this sense include sporting, cultural and business-oriented events, such as trade fairs. The event may be as brief as a few hours or last for a season and may take place on a single site or a range of sites. Events can be organised by government agencies, non-profit organisations or commercial businesses. Event organisation includes the production and marketing of the event itself and the coordination of a range of suppliers, including (in various proportions) venues, entertainment, security, accommodation and transport.
The production of events is not recognised as an industry in standard industrial classifications. If anything, it would be a subsector of tourism, which again is not recognised as an industry in the standard classifications, although the Australian Bureau of Statistics (ABS) has a definition of the sector and from time to time releases statistics about it. Persons employed putting on events may be classified as working in sport, the arts, business services, religion or,if all else fails, in entertainment not elsewhere classified.
An important distinction is that between events themselves and associated economic activity. Although accommodation, meals and transport may be arranged as part of an event, they are provided by recognisable related industries. There is also a distinction between events and event venues, sometimes expressed as arm’s length transactions (the event organiser hires the venue). However, events are frequently organised by venue operators while, in the opposite direction, organisations that are primarily event promoters may branch out into venue ownership. In the closely-related field of tourist attractions, venue and events may coalesce.
The predilection for user charges
Economic theory distinguishes between excludable and non-excludable events. The latter, exemplified by street parades and fireworks displays, are conducted in public spaces where it is impractical to charge for entry and where, indeed, some of the attendees may be unwilling witnesses rather than beneficiaries. Because it is not possible to charge attendees directly, the source of finance preferred in neoclassical economics, user charges, is not available and non-excludable events must, therefore, be financed from sources such as public culture and recreation budgets, business advertising budgets or the demonstration budgets of groups wishing to apply political pressure. Neoclassical economics allows that where governments finance a non-excludable event, they cannot be faulted for failure to impose user charges, although (as with all government expenditure) their priorities may be criticised.
Excludable events are those where it is practical to charge entry fees. The neoclassical principle is that both the event and the venue should be financed from user charges or, if available, by the use of voluntary labour. However, it is common practice for governments to contribute to the staging of excludable public events, sometimes by way of direct subventions, sometimes through more limited assistance (e.g. for marketing) and quite frequently by contributing to the infrastructure cost of venues. These contributions are contrary to neoclassical principle, but have been defended in two major ways.
A government may decide that an excludable event has such merit as a community occasionthat entry should be free, an approach often taken for events with educational, cultural or commemorative content. Short of this, it may decide that the event should be available at subsidised prices, either through a subsidy paid to the event provider or indirectly through venue subsidisation. A great many arts events fall in this category.
It may be argued that the event generates spin-offs for businesses that justify the subvention. In the extreme case, it may even be argued that the addition to business activity generates increased tax revenue such that the subsidising government actually gets its money back.
The first of these arguments is not particularly amenable to economic analysis because the benefits can only be measured indirectly. This is the domain of cost–benefit analysis, which specialises in the imputation of market-derived values to costs and benefits that are not directly valued by market transactions. Sometimes reasonably satisfactory surrogates are available, but sometimes they are not: for example, it has proved very difficult to put a market value on biodiversity. Where such values are important, difficult decisions have to be made as to how to evaluate mixed bags of market, near-market and distant-from-market values. In some studies, particularly those with environmental costs and benefits, NIEIR has included non-market costs and benefits, but other studies have concentrated on the second, purely economic effects of events. This may be termed economic impact analysis, as distinct from cost– benefit analysis, and concentrates on effects on employment and incomes.
Approaches to economic impact assessment
The economic impact of subventions to events can be estimated by economic modelling. Assessment using neoclassical models produces predictable results: events are assumed to absorb resources that would otherwise be employed in other types of production and, therefore, have negligible overall effects. Given the underlying assumption that the economy is subdivided, without remainder, into a series of perfectly-competitive price-equilibriated markets and is therefore fully employed, this result is inevitable. There is no need to have recourse to models if these assumptions are made, except to identify which areas of production contract in order to expand the production of events.
The position is rather different if the industries impacted by the event have spare capacity. This requires that markets be imperfectly competitive, including that production can be marginally increased without affecting price. A typical case would be a chef who does not increase prices just because his or her restaurant achieves 95 per cent occupancy for a night or two. Whether or not production can be increased without affecting prices is a practical matter, and cannot be solved by assumption. NIEIR has frequently been contracted to estimate the economic impact of events where it can be argued that the concerned industries have spare capacity and production is, for a range of capacity utilisation, demand-determined. The institute has evaluated both events where the existence of spin-off benefits is the primary defence of a subsidy and events whose primary justification is cultural or educational but where spin-off benefits are welcome and help to justify the subvention. Impacts may be calculated in advance (as part of advice as to whether the event should be supported) or in arrears (as part of an audit process and also for reference in deciding future applications). Either way, the calculation requires a comparison of two scenarios: a scenario in which the event is held (in the ex-post case what actually happened) and a scenario in which the event is not held. The impact of the event is estimated by differencing the scenarios.
The construction of formal scenarios requires both geographic and time definitions. These are adopted primarily according to client requirements but might also reflect data availability. The geographic area impacted may be limited to the immediate surrounds of the venue(s) but may be expanded to include their Local Government Area (LGA), the state or the whole country, while time definitions may include the run-up to the event, the event itself and its lasting effects. NIEIR can match these definitions to its datasets and models and, hence, provide assessments at the LGA, state and national levels (the sub-LGA level generally requires additional data collection) and for time periods defined by quarters (with the peak day or week a possibility subject to additional data collection).
Because the argument for subsidisation from spin-offs is expressed in economic terms, it is appropriate to measure the impact of the event by the addition to gross income obtained by differencing a scenario that includes the event and one which does not. (Gross income, in this context, means cash income before deduction of depreciation.) Some clients are also very interested in benefits in terms of employment generation and governments are sometimes interested in estimates of the amount of additional tax revenue generated by the event. The benefits thus calculated can be related to the size of the subvention and different subventions can then be compared for effectiveness and (with a little further modelling) can also be compared with the benefits of equivalent tax cuts.
Measuring benefits in terms of changes to income and employment does not provide a full cost–benefit analysis. Two elements are missing:
- assessment of additional external benefits, such as the educational value of the event or environmental benefits; and
- assessment of additional external costs, such as environmental costs and alternative leisure foregone to participate in, or work for, the event.
The simplest approach to these costs and benefits is to limit assessment to market values. This approach is regularly accused of ignoring much that is relevant. However, attempts to be all-inclusive can end up giving excessive weight to shadowy and elusive effects. NIEIR prefers to err on the market-value side but includes major external costs and benefits where it assesses them as relevant and assessable.
In evaluating events, it is not common to spend much effort on environmental effects. This is because few events have the primary purpose of generating environmental benefits and for most of them environmental costs are incidental. There is a substantial literature on environmental costs and benefits and defensible conventional values are available for many environmental costs, such as noise and carbon emissions. If these are considered significant they can be included in the assessment.
When events have educational value, it is theoretically desirable to include a value for this benefit. Various methodologies have been suggested, such as estimates of the cost of providing the same education by alternate means or estimates of the capital value of the addition to earning power resulting from the education. Educational and entertainment values are sometimes hard to distinguish, and attempts have been made to tie them down by surveys of willingness to pay over and above the actual admission charge and documentation of how much people have spent travelling to the event. These types of evaluation can quickly become subjective. Similar arguments apply to health effects.
A category of cost sometimes put forward is the cost of leisure foregone by people who work as a result of the event, to which one might add alternative leisure foregone by those who attend the event. As with all allocations of time, the basic point is simple: if people didn’t do one thing, they would do something else. The problem is that evaluation is necessarily subjective. Take the case of those who forgo leisure to work. At one extreme, the foregone leisure may have been, for example, a family occasion that the worker would much rather have attended but forgoes in order to retain a reputation for reliability with his or her employer. At the other extreme, the foregone ‘leisure’ may have been moping around for want of anything better to do or money to do it with. In this latter case, hours worked are not a cost but rather bring the benefit of approved social participation. The cost of ‘reservation wages’ is further discussed below.
Event assessment methodology
A decision to concentrate on income generation as the measurable benefit of subventions to events has the great advantage of bringing assessment into the realm covered by the National Accounts and, hence, of economic models based on the National Accounts. Thus, additional income resolves into additions to wage and salary income and gross profits within the selected geographic area. Remembering that the income and expenditure sides of the National Accounts are, by definition, equal in aggregate, the increases in gross income are conveniently measured as net additions to expenditure. There are two dimensions to this: direct expenditures that would not have taken place in the absence of the event (with an offset for direct expenditures curtailed by the event) and indirect expenditures that arise as a multiplier result of the direct expenditures.
Direct expenditures for the geographic area that includes the event venue comprise, primarily, expenditures to the area by visitors who came to attend the event and also, pro rata, expenditures by people who visited partly because of the event and partly for other reasons. There may be other additions: for example, a trade fair will generally receive additional income from the local expenditures of out-of-region exhibitors and may generate income from additional sales of local products, while the accommodation and other expenditures of a visiting sporting team, its officials and visiting media may be offset against the share of the gate revenue that the team takes home.
Sources of information on the direct impact vary from event to event. Subject to judgements as to relevance, NIEIR has used the following sources.
Surveys of attendees are conducted face-to-face at the event. All attendees are asked their age and gender and, depending on the study, they may be asked about their marital status, number of dependent children, occupation, household income and ticket type. Where events run for multiple days or cover multiple venues, the questionnaires cover days and places of attendance. After asking normal place of residence, local attendees (those who live in the geographic area or region defined for impact assessment) are asked their expenditure at and associated with the event, whether they would have travelled elsewhere to attend a similar event had the event not been held locally and whether they ran down savings to attend. Visitors from outside the region are asked whether the event is a major or merely contributing reason for their visit, their expenditure associated with the event (including its broad composition) and the means of transport and type of accommodation employed. Data are also collected on satisfaction with the event and the likelihood of attendance at subsequent similar events, but these data are for market research purposes and do not affect the assessment of benefits.
Similar data may be collected from event participants themselves. At trade fairs these include exhibitors and buyers, at sporting events the competitors and visiting officials and at cultural events the performers. However, for sporting and cultural events, information from the organisers may be substituted for direct interviews with participants.
At trade fairs, locally-based exhibitors may be asked to estimate the value of additional sales generated at the event. It is admitted that such estimates are necessarily approximate, if only because sales concluded after the event can only be anticipated.
Information provided by the venue owner and the event organiser and their subcontractors generally includes the event budget (particularly cash flows into and out of the region) and total patronage (so that survey data may be expanded to total coverage).
Interviewers are supervised and endeavour to question a random selection of attendees. Participant data may, however, be collected by distribution and later collection of forms. Data are compared to the results for other events as a check on validity and also to dimension particular events in relation to others in the same region.
Although local visitors are asked various questions for market research purposes, their only contribution to the economic analysis is the amount they spend financed by reductions in savings or by forgoing travel outside the region. All other spending is, by their own admission, financed at the expense of other local expenditures and, hence, does not alter demand within the region.
Greater attention is paid to visitors from outside the region. The number of net additional visitors from outside the region (typically interstate and overseas) who would not have come apart from the event is estimated, as is the number of visitors who extended their visit because of the event. The resulting addition to visitor days is estimated along with the number of bed-days of additional accommodation, by accommodation type. (Overseas visitors tend to stay in hotels or backpacker hostels, but interstate visitors frequently stay with relatives or in a caravan.) Crucially, an estimate is made of additional expenditure in the region, divided into: event entry fees and other revenue to the event organiser or venue; meals, food and drinks; entertainment; accommodation; transport; additional sales by local businesses resulting from trade fair or similar exposure (if any); and other.
The estimate thus prepared may be termed the addition to the exports of the region promoting the event. The estimates will vary according to geographic definition: as the geographic boundary is broadened from local through state to national, a higher proportion of expenditure becomes local and exports diminish.
None of this has proved particularly controversial. It is the next step that brings out the difference between neoclassical and post-Keynesian approaches.
Neoclassical assumptions to minimise assessed impact
We must first dispose of the neoclassical trump card, known technically as the reservation wage. The underlying assumption is full employment, meaning that all people who want paid employment are at work for precisely their desired number of hours. In these circumstances, increases in employment are not possible without increases in wage rates. People who take on work as a result of the increase in wage rates are accordingly giving up leisure to the value of the previous wage rate (because by the full employment assumption they would have worked at the previous wage had they not valued their leisure more highly). Therefore, the benefits of any additional employment are negligible. However, it has been noted that the value of leisure foregone is highly subjective and can even be negative, and will argue below that the full employment assumption is questionable. Reservation wages are irrelevant when the focus is on income and employment, and even in a strict cost–benefit context their value is highly debatable.
We now turn to the main point at issue, which is the valuation of expenditures curtailed by the event: expenditures which would have been made in the absence of the event but are not made due to the event. As already noted, the neoclassical analysis depends on the assumption of full employment and means that curtailed expenditures are similar to event-generated expenditures. Two variants of the neoclassical story have been proposed, both of which depend on labour being fully employed.
One story runs as follows. An event transfers expenditure from the regions of visitor origin to the region holding the event. In all regions, labour is fully employed. Increased expenditure in the region holding the event causes wage rates there to rise, while reduced expenditure in the regions of visitor origin causes wage rates there to fall. These changes in wage rates precipitate a flow of labour from the regions of origin to the region of the event. The transfer of expenditure is matched by a transfer of labour. This prediction is easily tested in practice. For no event that NIEIR has examined has there been any evidence that the required changes to wage rates have taken place or that the event has been staffed (either directly or indirectly) by labour migrating from the visitors’ home states.
A more plausible version of the neoclassical story runs as follows. As in the first neoclassical story, visitor spending raises the demand for labour in the region that holds the event and workers can only be attracted to provide for the demand by offering increases in wages. However, in this story, the response to the increased wages comes not from different geographical regions, but from the effect of the increased wage rates in reducing profitability in all other local industries. These industries sack employees who promptly transfer to the industries serving the increased visitor demand. The net effect on employment and the total value of output depends on price responses, but it is normally claimed that they more or less balance out: the increase in production to meet the increase in visitor demand is countered by a decrease in local production. Thus, a subvention to an event, even when it increases income earned from outside the region, does not increase net income except perhaps marginally where the price responses fail to balance out.
In theory, similar reasoning could be applied to capital capacity. Neoclassical economics includes the concept of the ‘short run’, defined as a period in which capital capacity cannot be altered. Because events are ephemeral, this would appear to be the relevant assumption. If capital capacity is fully employed, no changes in real output can take place in the short run and the capital capacity required to service visitor expenditure can only be made available by shifting it from alternative production, the same result as for labour.
The neoclassical analysis depends heavily on two assumptions:
- labour supply is very nearly fixed, although labour can be found to service additional demand at the cost of transferring it from alternative production and/or from valued leisure; and
- capital is fixed, although capital can be found to service additional demand at the cost of transferring it from alternative production (although this is the strict neoclassical position, some analysts assume that capital utilisation can vary and put the weight of their analysis on the labour supply assumption).
Whether or not capital capacity is assumed to be fixed, the operative assumption is that marginal costs are rising and translate into rising supply curves, not only in the industries serving visitor demand but generally. This assumption is made on purely theoretical grounds, without any investigation of labour market conditions applying in the area affected by the event at the time of the event. NIEIR does not contend that this assumption is always irrelevant: there can be times and places where additions to demand in one industry raise wage rates generally, and, more frequently, where they raise wage rates for particular skills; similarly, there can be times and places where additions to demand in one industry run slap into infrastructure or other capacity constraints. For example, mining booms raise wage rates generally in the remote mining regions and specifically for particular mining-related and construction skills, where they increase wage rates not only in the boom zones but across whole countries and, indeed, worldwide. Mining booms also crash against capital capacity constraints in the mining industry and for related transport, energy and water infrastructure. But is this universally true for events?
The obvious capacity constraint affecting events concerns venues. However, unlike mining booms, events are planned in advance with meticulous attention to venue availability. Only in the case of very large events, such as the Olympic Games, is venue availability a potentially limiting factor. Impact analysis for very large events is complicated by the need to evaluate the costs of the resulting planned construction program and to assess the benefits of the resulting assets after the event has taken place. Most events that NIEIR has been called upon to evaluate utilise existing venues. Where this is the case, the opportunity cost of the venue is reasonably represented by arm’s length venue hire charges. NIEIR includes these with the costs of staging the event.
Having dealt with venue and staging costs, the major question concerns the capacity of the local economy to service the addition to visitor expenditure, where visitors are defined as people coming from outside the region in which the venue is located. Judging by the composition of additions to demand revealed by the surveys conducted by NIEIR at events, these additions to demand chiefly concern accommodation, restaurants and other eateries, entertainment (not only the entertainment provided by the event itself) and transport. Except for the additions to transport demand, the additions are concentrated near the event venue.
The additions to expenditure on entertainment and private motor vehicle transport caused by a typical event (in size up to a Commonwealth Games or a papal visit) are minor in relation to total demand, even at the LGA level, and are likely to be within existing capacity: a few extra seats are sold and the average delay in the queue to pay the cashier at the petrol station increases by a minute or less.
Events can impose significant demand on local public transport. However, the transport authorities have the great advantage of fairly accurate estimates of likely demand and have shown themselves capable of marshalling their resources to meet the demand. Similarly, large events generate surges in air traffic, but the airlines have sophisticated demand management systems in place. If a surge in traffic is anticipated, their first step is to withdraw discount fares and the second is to schedule additional flights. If these flights require aircraft placement flights in the contrary direction the airlines may offer discounted fares to take people away from the event rather than to it. Additional flights are usually provided from reserve capacity in the existing fleet, including flying at unattractive times of day, and are manned from existing personnel resources. The increase in the demand for air transport that results from an event does not, therefore, withdraw labour from other industries. However, there may be an offset to the increase in demand from attendees at events from two sources:
Travellers who would have visited the event location in the absence of the event but who are put off by the increased fares (a group that also includes travellers who are put off by increased accommodation charges). Not all of these potential visitors are completely put off; some (helpfully) reschedule.
Residents of the event location who are persuaded to leave the area, either by discount airfares or simply to avoid being at home when the event is held. Once again, these departures may be opportunistic and merely rescheduled departures that would have taken place in any case.
For events significant enough to affect airfares, NIEIR estimates the offset to demand.
A stronger case can be made that events strain capacity in the hospitality industries: accommodation, restaurants and related services. Any event that is worth assessing will cause an increase in local demand for hospitality services sufficient to raise the demand for labour in these industries. The additional demand is met from two main sources:
- the offer of additional hours of work, either to existing part-time employees or to persons otherwise not in employment, at standard rates of pay; and
- the offer of additional hours of work at enhanced rates of pay.
Where additional employment is taken, at standard rates of pay, by people who would otherwise be in paid work there is no reason why rates of pay in other industries should be affected or that labour should be withdrawn from these industries. Whether or not labour can be recruited to an event on this basis is a matter of practical observation. The evidence that it can be includes the following:
The hospitality industries rely heavily on part-time workers. There is ABS evidence that, for most Australian locations and for most times since the end of full employment in the 1970s, many part-time workers desire additional hours at the going rate of pay.
The hospitality industries also rely quite heavily on low-skilled and semi-skilled workers. There is ABS evidence that, for most Australian locations and for most times since the end of full employment in the 1970s, many persons are available who would be happy to work at the going rate of pay were more jobs available at these levels of skill. Such workers include many more than those who are unemployed as officially defined. The surveys confirm the presence of significant numbers of people who do not meet the official definition of unemployment in terms of work search but who would, nevertheless, happily take on paid employment were it offered to them. We may also note that Australian labour-force participation rates are below those of many OECD countries, which would indicate the presence of potential employees to take jobs made available.
This said, some employers may prefer to meet the extra demand by raising wage rates. The obvious strategy is to employ existing staff on overtime, which raises costs in the hospitality industries but does not threaten to withdraw labour from other industries. If the increased costs are passed on in terms of increased prices there may, however, be a small offset to real demand in the form of reduced patronage by local residents who have to pay more to eat out.
Finally, demand may be met by the neoclassical mechanism of offering higher wage rates to attract workers who are already employed. In view of the excess supply of low and medium skill workers, increased wage rates are likely to be offered only to skilled personnel such as chefs. The point here is that hospitality skills are industry-specific; an increase in wage rates for skilled hospitality personnel cannot be guaranteed to generalise across all industry as required by the neoclassical hypothesis. However, as with overtime, there is a possibility of increased prices in hospitality reducing the real incomes of local residents and, hence, their real expenditures.
It can be concluded that an increased demand for labour in hospitality is unlikely to generalise to general increases in wages affecting all industries unless the region concerned is closer to full employment than has been the case in most Australian regions over the past three decades.
The position is similar with capital capacity. Event organisers generally organise hospitality at the same time as they organise the event and hospitality providers have shown themselves capable of a flexible response, much like the public transport providers and airlines. Hospitality providers customarily work with a level of excess capacity and also use flexible pricing. This means that a lack of discount accommodation reduces casual visitation during the event, although some of this is rescheduled. NIEIR makes an allowance for visitors who would have visited but for the event, deducting their estimated expenditures from those of the visitors who come because of the event.
For significant events occurring in the capital cities, it is possible to check whether an event was associated with price increases in the hospitality sector. The relevant indicator is the relationship between the price index for hospitality and the consumer price index in general. For several reasons it is expected that hospitality prices will trend upwards more rapidly than consumer prices: productivity increases in services are generally less than in manufactured goods; there is no competition from low-priced merchandise imports; and the hedonic price adjustment for information technology slows the rate of increase of the general index. Given these expectations, price increases for hospitality have been low and have not corresponded with surges in the demand for hospitality: indeed, the reverse has been the case, and periods of high increase in demand have tended to be associated with low increases in prices, as would occur if hospitality were an industry subject to increasing returns to scale.
It is argued by NIEIR that the supply of labour and capital to the hospitality industry in response to events has been cordoned off from the supply to other industries in most Australian locations at most times during the past three decades. This conclusion applies even in places like seasonal tourist resorts because the resorts draw on labour from the nearby cities or from sources such as young overseas tourists with 2-year work permits. If this is the case, additions to demand translate directly into increases in income. The offsets that neoclassical analysis assumes are, indeed, possible, but under typical conditions do not apply.
The stage is then set for multiplier effects. In traditional Keynesian fashion, an increase in demand generates an increase in incomes, the spending of which generates a further round of increased demand. The process continues subject to ‘leakages’, in the form of savings and capacity constraints, although we have checked the hospitality industry and noted that these are unlikely, and in conditions of generally less than full employment they are not particularly likely in other industries. However, it is admitted that skill supply and capital capacity constraints apply in some places at certain times, and when they do they limit multiplier effects as surely as leakage into imports.
NIEIR models these multiplier effects using:
- input–output tables, concentrating first on income generation in industries that supply the hospitality industry and then in second and subsequent rounds being more general;
- data on industry employee characteristics, to determine the kinds of household into which the additional incomes will flow, again concentrating on the hospitality industry in the first round;
- data on the characteristics of households receiving capital incomes;
- household expenditure patterns, to determine the directions of spending of households with increased incomes; and
- data on commuting patterns and trade matrices, to determine the locations of spending.
The results of this analysis are directly contrary to the neoclassical analysis. Instead of the initial increase in export demand being whittled away by reductions in production in other industries, it is expanded by increases in demand for the products of those industries. The result is that public subventions to events that attract visitor expenditure from outside a region generally benefit the region, the exception being where they hit capacity constraints. The returns to public expenditure can be attractively high, although there is no guarantee: NIEIR has come across events that have failed to attract non-local visitors and where the economic impact has been negative. Subventions to such events may still be justified on cultural or other grounds, but not on grounds that the events generate income or employment.
Should generally positive returns guarantee priority in public budgets?
The finding that, under present circumstances of less than full employment, events are often effective means of translating public funds into employment increases is good news for event organisers but less to the taste of those in Treasuries whose unenviable but worthy task is to balance the public budget and to ensure that public spending yields value for money. Rather than rely on spurious neoclassical arguments that event subsidies cannot, by assumption, yield positive economic impacts, Treasuries would do better to rely on the following arguments.
First, the assessment that a subvention to an event yields increases in income and employment does not give the event an absolute claim on government funds. Technically speaking, the rate of return on each event should be compared with the rates of return on the following: subventions to other events; other government expenditures; tax cuts; and increasing government saving. These comparisons depend on the macroeconomic circumstances of the day and also rely on similar assessments being available for alternative government expenditures. Needless to say, this is a rather tall order, but the principle remains that governments have the unenviable task of deciding the relative urgency of different expenditures. Although rates of return are helpful in determining priorities, they do not do away with the necessity to choose.
A second argument sees event promotion as a zero-sum competitive game between the Australian states. According to this view, event promotion is as pernicious as local preference in awarding manufacturing contracts and it simply encourages inefficiency. This argument is incomplete in that it fails to recognise that events increase national exports as well as interstate exports; competition by states to stage events is not a zero-sum game. Competition between places goes beyond competition for the tourist dollar; it is, in part, competition for identity and recognition. Success in such competition brings intangible benefits to residents, which have not so far been mentioned in the analysis. Turning to a visitor point of view, competition between cities and towns increases the range of visitor experiences available: efficiency arguments that assume that events are homogenous miss the point of this type of competition. Also note that the Australian states not only compete for interstate tourism, but combine to promote international tourism. There may be a point where competitive tourism promotion by the Australian incurs diminishing returns but it is unlikely that this point lies at zero expenditure.
Third, events are not the only ‘industry’ to benefit from public support. For example, subventions for events and capital investment in venues are small compared to government expenditures on road construction and maintenance. Both provide direct consumption benefits to citizens and both generate benefits to private businesses. This is the exact point of calculating the income generation benefits from different forms of public expenditure. Instead of assuming benefits away, as they do when they adopt the neoclassical assumptions, Treasuries should compare the benefits of alternative spending patterns and adjust their spending accordingly.
Fourth, the benefits from event subventions need to be placed in the context of general macroeconomic policy, which in Australia for several decades has favoured the generation of demand by the encouragement of bank lending, and, more specifically, the encouragement of mortgage lending to households. The Commonwealth was so anxious to pursue this policy that it was unwilling to increase real taxes and limited government borrowing, thus limiting the flow of funds available for event subventions and for all other forms of government spending. The tragedy pointed out in recent State of the Regions reports was that the policy of high mortgage lending has failed to generate either full employment or affordable housing, the ultimate culprit being the constrained supply of accessible urban land, which is partly due to the constraints to government investment required to create an enhanced supply.
This brings us back to the question of constraints more generally. While it is argued here that there are no constraints that prevent the operation of income multipliers arising from visitor expenditure, there may be constraints at a more general macroeconomic level.
As described in the National Economic Review as far back as June 1987, Australia suffers from a number of serious macroeconomic constraints:
- a public sector constraint – broadly, the unwillingness to pay sufficient taxes to finance desirable public expenditure, particularly public investment;
- various capital capacity constraints, including shortages of highly specialised skills and an inability to make the public investments required to alleviate these shortages, and various infrastructure inadequacies arising again from low public investment; and
- a balance of payments constraint, arising from reliance on an exchange rate that has failed to settle at rates justified by the economic fundamentals coupled with industry policies that pay insufficient attention to investment in export industries – not merely the mining industry, but industries that are capable of generating jobs that match the skills available.
In view of these constraints, the high benefit–cost ratios commonly observed for subventions to events should be assessed in the context of a long-term strategy for Australian economic development. Event assessments are but a tiny part of the analysis required to develop and implement such a strategy.
Neoclassical economics purports to show that economic policy should concentrate on leaving decisions to markets. As applied to the analysis of events, the neoclassical policy recommendation is for reliance on user charges except where there are non-economic reasons for free or subsidised provision.
NIEIR’s analysis of event provision shows that this recommendation relies on assumptions that have not applied in most parts of Australia over the past 30 years. Instead, subventions to events that increase tourist visitation can increase incomes and employment. Thus, the case for subventions becomes one of priority against other expenditures, preferably exercised in the context of a coherent strategy for the future.