One major finding regarding the overall cost of elections is the importance of the type of democracy environment (i.e. stable, transitional and post-conflict) in determining both the kind and amount of electoral expenses. The conclusions from previous research, have been validated in the current study and are described according to the two categories below:
A very significant factor in explaining cost variations is duration of previous experience with multi-party elections. Significant cost differences exist between routine elections in stable democracies, elections in transitional democracies, and elections during special peacekeeping operations. In countries with longer multi-party democratic experience, elections are consistently less costly than in countries where such elections constitute a new undertaking. This trend cuts across regions, levels of economic development, and even interruptions of electoral practice by military breakdowns. Low electoral costs, approximately $1 to $3 per elector, tend to manifest in countries with longer electoral experience: the United States and most Western European countries; Chile ($1.2), Costa Rica ($1.8), and Brazil ($2.3) in Latin America; Benin ($1.6), Botswana ($2.7), Ghana ($0.7), and Senegal ($1.2) in Africa; India ($1) and Pakistan ($0.5) in Asia; and Australia ($3.2).
In most countries that have less multi-party electoral experience, costs tend to be higher, even taking into consideration elections that have taken place as part of peacekeeping operations, where the cost per elector is highest: Mexico ($5.9), El Salvador ($4.1) and Paraguay ($3.7) can be mentioned in Latin America; Lesotho ($6.9), Liberia ($6.1) and Uganda ($3.7) in Africa; and Russia ($7.5) in Eastern Europe. Thus, duration of electoral practice is in itself a cost-reducing mechanism, perhaps the most important during the stage of democratic consolidation. Since a longer-term perspective is by definition difficult when assessing election costs in new democracies, the above findings offer strong support for the claim that efforts at capacity building in electoral administrations are probably cost-effective in the longer term. These findings also support the idea that establishing and consolidating a permanent electoral administration as the repository for managerial capacity development with regard to elections—within both the political and the administrative systems—is a cost-effective practice.
As might well be expected, elections held as part of broader and longer-lasting peacekeeping operations are the costliest of all. Nicaragua in 1990 ($11.8 per elector), Angola in 1992 ($22), Cambodia in 1993 ($45.5), Mozambique in 1994 ($10.2), Palestinian Territories in 1996 ($9), and Bosnia-Herzegovina under the Dayton Accords ($8) are cases in point. This is not to say that a cost-effective approach cannot or should not be used for special operations, but that it would function to a much more limited extent than in simple transitional electoral politics or, indeed, in routine periodic elections. In the Cambodian case, in which donors subsidized both elections, it would be hard to demonstrate that the high-cost elections in 1993 (at $45 per elector) were better organized or produced a more positive political outcome than did those of 1998, which were run at costs closer to the standard of the politics of democratization ($5). Somewhat less dramatically, both Nicaragua and El Salvador also demonstrate that second elections after peacekeeping operations can be run significantly less expensively: costs dropped from $11.8 in 1990 to $7.5 in 1996 in Nicaragua, and from $4.1 in 1994 to $3.1 in 1997 in El Salvador. Consequently, elections as part of special peace-making and peacekeeping operations should be considered separately for both analytical and strategic policy purposes. (López-Pintor, 2000, 76-77).
According to the research conducted for the CORE Project, elections cost more than $20 per elector in Afghanistan (2004), $5 in Guatemala (2004), $2 in Cambodia (2003), $4 in Spain (2004), and $2 in Sweden (2004). The projected per-elector cost for the 2005 election in Haiti is $11. Costs in Guatemala and Cambodia are similar to those in Spain and Sweden, respectively, but these similarities should be considered in light of the fact that the former two countries are among the poorest in the world in terms of per capita income, while the latter two are among the richest. Thus the elections in Cambodia and Guatemala can be viewed as being much more costly.
A corollary to these findings is that the integrity costs of elections are reduced by efforts and investments to improve peace, security and national reconstruction, such as disarmament, demobilization, integration to civilian life, inter-ethnic reconciliation, and infrastructure development. Other electoral costs may remain constant or even increase (e.g., personnel, high technology), but integrity costs will certainly decline with democratic progress. Within the realm of core costs, progress in building state apparatuses would normally imply that certain electoral costs may be reduced or shared within the ordinary budget of other public agencies different from the electoral administration (e.g., civil registries, postal services and police). As national security, transport and communication infrastructures are of paramount importance in determining integrity costs, so is state building with regard to core electoral costs.
Democratic consolidation
Core costs as a whole—especially in the areas of personnel and advanced technology—tend to increase rather than decrease independently according to the degree of democratic consolidation. One main cause of this in emerging democracies is the sheer institutionalization of a permanent professional electoral administration, which in most countries is a bureaucratic organization in the form of an electoral commission independent of the executive branch. A second set of causes stems from the complexities of the political and party systems in countries with federal, state and regionalized electorates, all of which may run different types of elections under separate schedules; when high demands for voter information persist (i.e., use of different languages); and when elections require special voter assistance (i.e. external voting, proxy voting). A third cause is the increasing use of new technologies, which may include computerization of office work, establishment and update of permanent computerized voter lists, quick electronic transmission of results on election evening, and introduction of electronic voting. All of these activities are expensive.
In the specific scenario of emerging democracies, once peace is achieved and reconciliation fares well, dramatic cost decreases may be expected—though not under all circumstances (e.g., Ecuador, Kosovo and Nicaragua). The case of Cambodia illustrates the decreasing cost trend and the fact that after a given point, no further decrease can be expected. The substantial reduction of costs achieved by the National Election Committee (NEC) from 1998 to 2003 can be explained by a series of factors that are not all technical in nature. At the policy level and as a matter of pride, the Government of Cambodia does not want to rely heavily on international assistance for future elections. Eventually, the greater part of electoral costs will come from the government’s coffers. This, in itself, is a strong incentive to reduce costs. Electoral assistance dropped substantially, from almost $19 million in 1998 to $10 million in 2002 to $6 million in 2003.
The 2002 reform of the legal framework in Cambodia was undertaken with the double objective of improving and reducing the costs of the electoral machinery. For instance, a permanent voter registry was introduced, and voter registration was delegated to local administrations. In addition, expensive procedures were abandoned or replaced with cheaper alternatives. For example, the practice of issuing a new voter identity card (with photograph) for each election was replaced by the use of other identification documents, such as a new national identity card. The NEC was forced to rationalize its internal practices and reduce the number of departments. The chairman of the NEC has said he believes that electoral costs in Cambodia should not vary widely in future elections; if anything, he said, the cost of materiing factors to larger electoral budgets in the future may include capital expenditures to replace aging vehicles, computers and other electoral items, and the necessity to provide for an ever-increasing voter population.
In Guatemala, an assessment of election budgeting, funding and cost management in recent years found that election costs have consistently increased overall and in almost every single area. First, the 2003 election budget was more than twice that of previous elections after including foreign aid, but still around the average for Latin America. A similar pattern observed in Guatemala at the time of these elections had been seen previously in elections in Nicaragua in 2001 and Ecuador in 2002. In all three cases these were third-generation elections which took place after acute social conflicts, and each poll turned out to be more expensive and more poorly organized than the previous national election. Given that a substantial part of Guatemala’s electoral budget was funded through foreign aid, this finding should be taken as a warning to the international community to closely monitor election spending.
Second, the most expensive items in Guatemala were salaries and per diem given to temporarily hired staff and polling station workers, although this cost has remained relatively constant over the last five years at 125 quetzals ($15) for polling station officers— an amount that many poll workers did not consider equitable. This expenditure was followed closely by the cost for voter list updates, although it should be noted that updating of voter lists and rearrangement of the polling stations accordingly was basically covered by external aid. Cost increases in the official national electoral budget basically involved larger expenses for voter education and an increase in the number of polling stations to 8,885. Over a five-year period, voter information and education had the largest real growth in costs, most of which stemmed from an effort to reach rural and indigenous populations in their own languages (half of the Guatemalan population is of Mayan descent and does not speak Spanish). There is no single item for which a cost decrease has been recorded.
Next: Electoral budget histories in 1990s
