Introduction

Throughout history, public participation in the various economic sectors has been a bone of contention argued over by the supporters of public ownership and the free market. Despite of this, public participation in some specific economic sectors such as health has become a fundamental premise in developed States, and this sparks further debate on the suitability of the public sector’s breadth of intervention or investment.

In the case of public administrations, once public resources have been allocated, the estimated amount of revenue places a limit on public expenditure with no further expenditure having to be incurred for income to be raised.

This limitation has dual implications. First, the expenses that are to be met need to be prioritized, with the available resources allocated according to the needs determined to be the most important. Second, the decision to use more or fewer resources on a specific expense implies limiting to a greater or lesser extent the resources that can be devoted to other expenses. All this makes effective and efficient management of public resources essential.

The Spanish health system is considered to be the most efficient in Europe and the third most efficient in the world after Hong Kong and Singapore (Bloomberg Report 2018). The Bloomberg Report measures the efficiency of health systems by life expectancy, per capita spending on health and the percentage of GDP spent on health.

According to the Euro Health Consumer Index report1, Spanish health is in nineteenth place with 698 points, behind Serbia and Slovakia. The report highlights the fact that the Spanish health system is "very regionally decentralized".

The Spanish health system is decentralized. It is implemented by the country’s autonomous regions—the seventeen autonomous communities (CAs), all of which operate their own independently-managed health services, which can, therefore, be compared and improved.

The purpose of this study is to detect the Spanish health services that are efficient and those that are not. Measures are then proposed that can be implemented by any health services that are inefficient for them to be able to achieve a level of efficiency and, consequently, help improve the Spanish health system overall.

Methodology

Para consultar este apartado, ver el artículo en PDF / To consult this section, see the article in PDF

Results

According to the models based on studies (Sueyoshi et al., 2012) were applied to the variables explained. The corresponding data are given in Table 1.

Table 1.

Original data used in this study.

The results obtained from applying the model (Sueyoshi et al., 2012) can be examined in detail in annexes I and II, note that the programming of the DEA models is done using linear programming in Mathematica 12.3. The first column of both tables indicates natural or managerial efficiency, respectively, taking a value between 0 and 1 being 1 efficiency. Columns I1, I2, I3 deviations indicate the increase in the budget required to achieve efficiency (value 1 in the first column), in each of the inputs studied. Columns O1, O2 deviations indicate the increase or decrease in the degree of satisfaction and the mortality rate, respectively, when achieving efficiency.

Linked to the variation in resources invested in the various health systems, the effect that this variation has on the results has to be taken into account. For this, the Return to Scale (RTS) should be observed that indicate the effect that any variation in the Inputs has on the Output-G (good Outputs). This rate could be increasing, constant, or decreasing, which implies that any variation in the Inputs causes a variation in the Output-G that is proportionally greater, proportionally the same, or proportionally smaller, respectively.

The Damage to Scale (DTS) that link the variation in Inputs to the variability of the Output-B (bad Outputs) also need to be considered. As with the RTS, any variation in the Inputs causes variations in the Output-B that are proportionally greater, the same or smaller.

Annex I (column RTS and DTS) show the increases that these health services require to their health care budgets. These increases would impact on their Outputs and ensure that their production functions are in line with the set of health services that are part of the efficiency border that the remaining health services should target.

Sueyoshi et al. (2012) address environmental issues. In such a context, any increase in Inputs inherently worsens the results, i.e., there is an increase in bad Outputs and a decrease in good Outputs. This study’s context is different. Increasing the Inputs intrinsically raises the good Outputs and reduces the bad Outputs.

The efficiencies calculated with this model should, therefore, be interpreted the other way round, which means that environmental studies dealt with in natural efficiency health system studies should be interpreted as managerial efficiency and vice versa.

Based on the above data and following the grouping taken from the combination of rescaled difference clusters of natural efficiency, health services can be classified into the following groups:

Table 2.

Classification of health services by cluster analysis

Table 2. cont.

Classification of health services by cluster analysis

As can be observed, applying the two models, natural and managerial, classifies health services into four groups, which will be analyzed in the following.

Graphic 1.

Classification of health services by cluster analysis

Group 1: Health Services with high natural efficiency

This group is formed of the health services with either natural efficiency or with minimum health inefficiency as, in the worst case, the service’s inefficiency does not reach 1.405382% (Murcia case, see Annex I). These are the health services of the Cantabria, Catalonia, Madrid, Navarre, Basque Country, La Rioja, Castile-Leon, Aragon, Asturias and Murcia CAs.

The first five of these health services are naturally efficient, which implies that they have sufficient resources that are being well-managed.

Also, the Murcia CA health service does not require any increase to its budget and its budget does not need to be redistributed (the value in “I1 deviations”, “I2 deviations” and “I3 deviations” are 0). However, its available resources should be better exploited for the results to give a 1.405341% (annex I, column “01 deviations”) boost to civic satisfaction and lower the mortality rate by 6.68588% (annex I, column “02 deviations”).

The remaining health services are also approaching efficiency, but all require a minimum increase to their budgets, which should have a natural knock-on effect that improves their efficiency. This improvement would result in an increase in Output-G and a decrease in Output-B.

Group 2: Health services with mid-to-high natural efficiency

This group is composed of the Balearic Islands health service alone. The inefficiency of this health service does not reach 0.05. However, it is beginning to deviate from the efficiency of the previous group’s health services.

This health service requires a budget increase of around 190€/inhab. (Annex I, columns “I1 deviations” + “I2 deviations”), mostly for health expenditure directly managed by the public service and only a small part allocated to training. If the current management of public resources remains the same, this budget increase would result in a small 0.3186 (annex I, column “01 deviations”) improvement in the degree of civic satisfaction and a general 26.8559 (annex I, column “02 deviations”) reduction in the age-adjusted mortality rate per 100,000 inhab.

It must be borne in mind that RTS is decreasing, which implies that Output-G variability is less than proportional with respect to the increase in Inputs. This means that, in order to achieve efficiency, a more-than-proportional increase in inputs is required to achieve the desired improvement in the output that requires the increase. DTS is constant, which means that the impact on the improvement in the Output-B is proportional to the budget effort made and, consequently, also on the advantage achieved from this variable.

Group 3: Health services with mid-to-low natural efficiency

This group is formed of health services with a natural efficiency between 0.878165074 (Galicia case, annex I) and 0.905349683 (Castile-La Mancha case, annex I). These are the health services of Castile-La Mancha, Extremadura, Valencia, and Galicia. All these health services have constant RTS and DTS rates. This implies that the increase in resources allotted to their health systems managed both directly by the public service and set aside for training and external contracts with private companies results in a proportional improvement in the results in the form of both an increase in user satisfaction and a reduction in the mortality rate.

These increases in resources and their recommended distribution, which can be observed in detail in annex I, cause a natural improvement in both civic satisfaction with the various health systems and to the age-adjusted mortality rate per 100,000 inhab., and positions the health services of these CAs on the border marked by the production functions that operate efficiently naturally.

Group 4: Health services with low natural efficiency

The health services that form this group are those of Andalusia and the Canary Islands. These health services have an efficiency level of between 0.845352402 and 0.861361693 (see annex I) and require an increase in their budget effort. In the case of Andalusia, an increase of €420.50 (annex I, column “I1 deviations”) per protected citizen is required to health expenditure directly managed by the health service and €64.81 (annex I, column “I3 deviations”) to health expenditure for external contracts. The increase to the budget required to achieve efficiency in this Autonomous Community is significant (almost €500 per inhabitant).

Depending on the model applied, this increase to the budget will result in a 13.8636% (annex I, column “01 deviations”) rise in civic satisfaction with the health system and a 16.3069% (annex I, column “02 deviations”) reduction in the mortality rate.

The following table shows in graphic form the variation required in Inputs and Outputs (should any variation be necessary) for health efficiency to be achieved naturally.

None of the adjustments referred to in the previous paragraphs require any changes to their management at all but they do need the resources that they have been allocated to be better exploited with the use of the same procedures.

It must be borne in mind that the philosophy that governs the management of public budgets is totally different from the basis for their management in the private sector. In the public sector, administrations have resources that do not depend on expenditure, although these do limit them, i.e., the maximum that budgeted public expenditure can reach is the total amount of budgeted public revenue. This implies that an increase in the budget for any given expenditure causes a reduction in the budget for some other expenditure, which means that although resource management is fundamental in any organization, this is even more the case in the public sector.

For this reason, apart from seeking efficiency in this type of public organizations by way of a higher resource budget, it is desirable for their management to be improved through changes to management policies that lead to a simultaneous savings in resources and better results. This is the reason why this study groups health services together according to the criteria taken from the combination of rescaled difference clusters of managerial efficiency. This enables the health services to be classified into four groups, although this entails different implications:

Group 1: Health services with high managerial efficiency

This first group is the most populated. The health services that comprise this group are the Andalusia, Aragon, Balearic Islands, Cantabria, Castile-Leon, Castile-La Mancha, Catalonia, Madrid, Navarre and La Rioja CAs. These are all health services run with managerial efficiency, which means that the production functions of said health services define the efficiency border that the remaining DMUs must target to achieve efficiency. They all manage their resources efficiently, so no changes are required to their budget management procedures.

In the case of Andalusia, note that it is inefficient by taking into account the natural efficiency, while it is efficient if we take into account the managerial efficiency. This leads to an additional question being posed with regard to the congestion of resources. The law of diminishing marginal returns states that when a given level of resources is reached, excess resources lead to negative returns, i.e., any increase in Inputs can inherently cause a worsening of Outputs due to the management of the former being hampered. This is an issue that will be addressed in a forthcoming study.

Group 2: Health Services mid-to-high managerial efficiency

The second group is composed of health services that, despite not achieving managerial efficiency, have an inefficiency between 0.04277 (Basque country case, annex II) and 0.05246 (Valencia case, annex II). The group includes the health services of Asturias, Valencia, Murcia and the Basque Country. In most cases, for managerial efficiency to be achieved they need to use fewer resources while nonetheless improving the obtained results.

The biggest budget cutback should be made by the Basque Country health service, which needs to make an almost €270 (annex II, columns “I1 deviations” + “I2 deviations”) per person cut in its investment in health services. Most of this should be to direct management by the health service with a small reduction to the training of residents. In Asturias and Murcia, only a minimal cutback is required and in Valencia, there is no need for any cutback at all. Nevertheless, in cases where new, reduced budgets are necessary, the health services need to change their resource management policies to achieve better results, raise the level of civic satisfaction and reduce the mortality rate in their respective CAs.

In the specific case of the Basque Country, as the population of this CA was 1,049,555 inhabitants in 2016 (according to the INCLASS database [2019] from which all the other data have been taken), the savings made with a cutback of this type would be approx. €284 million, which could be allocated to some other public services.

Group 3: Health services with mid-to-low managerial efficiency

This group is formed of health services with managerial efficiency between 0.890822407 (Extremadura case, annex II) and 0.894502012 (Galicia case, annex II). In this case, they are the health services of the Galicia and Extremadura CAs. The reduction in the budget is minimal in both cases. In Extremadura, this could represent a saving of almost €27 million (537,391 inhabitants dispersed throughout the territory).

The challenge that this health services face is to improve their management processes, i.e., they have to achieve better results that improve the level of civic satisfaction between 0.7107 (Extremadura case, annex II, column “01 deviations”) and 0.7299 (Galicia case, annex II, column “01 deviations”) and reduce the mortality rate per 100,000 inhabitants by 48-52 persons, i.e., with the same resources but changing the way that they work.

Group 4: Health services with low managerial efficiency

Lastly, the health system with the poorest managerial efficiency is that of the Canary Islands CA. In this case, no cutback in the budget is required but the obtained results need to be improved with the resources that they have.

The following table gives a graphic representation of the variations required in inputs and outputs, where these are necessary, to achieve managerial efficiency in health.

Discussion

Good management of public resources is of vital importance for the proper running of the welfare State. The budget limitations to which the administrations are subjected imply that there is a need to prioritize the investment or expenditure of the resources that they have available. For this reason, this study analyzes the efficiency of the health systems dependent on the CAs in Spain and proposes corrective measures for the inefficient units to achieve efficiency.

The study is based on the application of the DEA model and natural and management efficiency of the health services dependent on the seventeen CAs into which Spain has its territorial organization decentralized. This study extends previous research that has demonstrated the adequacy of the DEA method and allowed researchers to propose corrective measures to achieve overall efficiency in inefficient health systems (Campos et al., 2016) as well as to look for efficiencies can be attained through natural and/or managerial adjustments by increasing the variables that are considered desirable (Campos et al., 2018) arriving at conclusions according to the search for the natural and managerial efficiency.

In the case of natural efficiency, any shortfall can be corrected by making available a higher budget that should be properly managed. This is a fundamental difference from the application made by other studies that apply DEA.

In the case of managerial efficiency, savings to the budget is the main priority. In the specific sector under study, a saving in any given expenditure means an opportunity to increase some other. Managerial efficiency implies that even when fewer resources are available, a change in the way that things are done means that better results can be obtained.

In short, all the health services that are not managerially efficient have to change the way that they operate. They must change their resource management policies in order to achieve better results with the same or a lower budget.

Specifically, as the data demonstrate that the lack of efficiency is to a great extent the result of mortality rates that must be reduced, the resources that are freed up could be allocated to putting preventive medicine policies into effect, which improves public health; restructuring the team of professionals, tasked with putting a greater effort into the early detection of illness; lower patient waiting times, which can be resolved with improvements to the management of operating rooms; and specialized centers with different schedules, which can bring a halt to delays in the detection of illnesses. More resources could be allotted to research, to improving technology, to talks given by experts on important and especially called-for subjects, to the redistribution of resident staff in line with agreed criteria, to the application of incentivizing policies that motivate residents and to improving labor conditions that enable the greater productivity of residents, etc.

Annexes

Annex I

Annex II

References

Adang, E. M., & Borm, G. F. (2007). Is there an association between economic performance and public satisfaction in health care?. The European Journal of Health Economics, 8(3), 279-285. https://doi.org/10.1007/s10198-007-0045-6

Alonso, J. M., Clifton, J., & Díaz-Fuentes, D. (2015). Did new public management matter? An empirical analysis of the outsourcing and decentralization effects on public sector size. Public Management Review, 17(5), 643-660. https://doi.org/10.1080/14719037.2013.822532

Begley, C. G., & Ioannidis, J. P. (2015). Reproducibility in science: improving the standard for basic and preclinical research. Circulation research, 116(1), 116-126. https://doi.org/10.1161/CIRCRESAHA.114.303819

Campos, M. S., Fernández-Montes, A., Gavilan, J. M., & Velasco, F. (2016). Public resource usage in health systems: a data envelopment analysis of the efficiency of health systems of autonomous communities in Spain. Public health, 138, 33-40. https://doi.org/10.1016/j.puhe.2016.03.003

Campos Lucena, M. S., Sanz Díaz, M. T., Velasco Morente, F., & Yniguez Ovando, R. (2018). The efficiency of health services in Spain based on data envelopment analysis (DEA). The singularity of health efficiency. International Journal of Collaborative Research on Internal Medicine & Public Health, 10 (1), 776-791. https://hdl.handle.net/11441/98523

Campos Lucena, M. S., & Velasco Morente, F. (2013). The efficiency of public television in Spain: application of the DEA measurement model. Revista de Ciencias Sociales, 19(2), 362-378. ISSN (Versión impresa): 1315-9518

Choi, T. M., Wallace, S. W., & Wang, Y. (2016). Risk management and coordination in service supply chains: information, logistics and outsourcing. Journal of the Operational Research Society, 67(2), 159-164. https://doi.org/10.1057/jors.2015.115

Clement, J. P., Valdmanis, V. G., Bazzoli, G. J., Zhao, M., & Chukmaitov, A. (2008). Is more better? An analysis of hospital outcomes and efficiency with a DEA model of output congestion. Health care management science, 11(1), 67-77. https://doi.org/10.1007/s10729-007-9025-8

Cooper, W. W., Seiford, L. M., & Tone, K. (2007). Data envelopment analysis: a comprehensive text with models, applications, references and DEA-solver software (Vol. 2, p. 489). Springer. ISBN: 978-0-306-47541-2

Donghua, L., Lan, Q., & Ying, B. (2008). The M & A Performance of the Listing Medical Companies in China-Based on DEA Analysis. Reform Strategy, 10, 054.

El-Noush, H., Silver, K. L., Pamba, A. O., & Singer, P. A. (2015). Innovating for women’s, children’s, and adolescents’ health. bmj, 351. https://doi.org/10.1136/bmj.h4151

Emrouznejad, A., Parker, B. R., & Tavares, G. (2008). Evaluation of research in efficiency and productivity: A survey and analysis of the first 30 years of scholarly literature in DEA. Socio-economic planning sciences, 42(3), 151-157. https://doi.org/10.1016/j.seps.2007.07.002

Euro Health Consumer Index (2018): 0 (Accessed June 26, 2019).

Expósito, A., & Velasco, F. (2020). Exploring environmental efficiency of the European agricultural sector in the use of mineral fertilizers. Journal of Cleaner Production, 253, 119971. https://doi.org/10.1016/j.jclepro.2020.119971

Farrell, M. J. (1957). The measurement of productive efficiency. Journal of the Royal Statistical Society: Series A (General), 120(3), 253-281. https://doi.org/10.2307/2343100

Girth, A. M., Hefetz, A., Johnston, J. M., & Warner, M. E. (2012). Outsourcing public service delivery: Management responses in noncompetitive markets. Public administration review, 72(6), 887-900. https://doi.org/10.1111/j.1540-6210.2012.02596.x

Glasgow, R. E., Vinson, C., Chambers, D., Khoury, M. J., Kaplan, R. M., & Hunter, C. (2012). National Institutes of Health approaches to dissemination and implementation science: current and future directions. American journal of public health, 102(7), 1274-1281. https://doi.org/10.2105/AJPH.2012.300755

Glover, F., & Sueyoshi, T. (2009). Contributions of Professor William W. Cooper in operations research and management science. European Journal of Operational Research, 197(1), 1-16. https://doi.org/10.1016/j.ejor.2008.08.011

Goddard, M., & Smith, P. C. (2003). Los indicadores de gestión en el sector público: fortaleza y limitaciones. Papeles de economía española, (95), 35-46.

González-Rodríguez, M. R., Velasco-Morente, F., & González-Abril, L. (2010). La eficiencia del sistema de protección social español en la reducción de la pobreza. Papeles de población, 16(64), 123-154.

Hadad, S., Hadad, Y., & Simon-Tuval, T. (2013). Determinants of healthcare system’s efficiency in OECD countries. The European journal of health economics, 14(2), 253-265. https://doi.org/10.1007/s10198-011-0366-3

Hartley, J. (2005). Innovation in governance and public services: Past and present. Public money and management, 25(1), 27-34. https://doi.org/10.1111/j.1467-9302.2005.00447.x.

Hodge, G. A., & Greve, C. (2007). Public–private partnerships: an international performance review. Public administration review, 67(3), 545-558. https://doi.org/10.1111/j.1540-6210.2007.00736.x

Ijiri, Y., & Sueyoshi, T. (2010). Accounting Essays by Professor William W. Cooper: Revisiting in Commemoration of his Ninety‐Fifth Birthday. Abacus, 46(4), 464-505. https://doi.org/10.1111/j.1467-6281.2010.00313.x

INCLASNS (2019), “Base de datos de Indicadores Clave del Servicio Nacional de Salud” del 2019 (Accessed February 5, 2019). http://www.msssi.gob.es/estadEstudios/estadisticas/sisInfSanSNS/inclasSNS_DB.htm.

Lowell, C. A. K., & Muñiz, M. (2003). Eficiencia y productividad en el sector público: temas dominantes en la literature. Papeles de Economía Española, 95.

McKee, M., Stuckler, D., & Basu, S. (2012). Where there is no health research: what can be done to fill the global gaps in health research?. PLoS medicine, 9(4), e1001209. https://doi.org/10.1371/journal.pmed.1001209

Moschuris, S. J., & Kondylis, M. N. (2006). Outsourcing in public hospitals: a Greek perspective. Journal of Health Organization and management. 20, (1), 4-14. https://doi.org/10.1108/14777260610656534

Sueyoshi, T., & Goto, M. (2011). Measurement of returns to scale and damages to scale for DEA-based operational and environmental assessment: how to manage desirable (good) and undesirable (bad) outputs?. European journal of operational research, 211(1), 76-89. https://doi.org/10.1016/j.ejor.2010.11.013

Sueyoshi, T., & Goto, M. (2012). Returns to scale and damages to scale on US fossil fuel power plants: Radial and non-radial approaches for DEA environmental assessment. Energy economics, 34(6), 2240-2259. https://doi.org/10.1016/j.eneco.2012.07.024

Sueyoshi, T., & Goto, M. (2013). A comparative study among fossil fuel power plants in PJM and California ISO by DEA environmental assessment. Energy Economics, 40, 130-145. https://doi.org/10.1016/j.eneco.2013.06.004

Sueyoshi, T., & Goto, M. (2014). Investment strategy for sustainable society by development of regional economies and prevention of industrial pollutions in Japanese manufacturing sectors. Energy Economics, 42, 299-312. https://doi.org/10.1016/j.eneco.2014.01.001

Thune, T., & Mina, A. (2016). Hospitals as innovators in the health-care system: A literature review and research agenda. Research Policy, 45(8), 1545-1557. https://doi.org/10.1016/j.respol.2016.03.010

Wiley, M. T., Rivas, R. L., & Hristidis, V. (2016). Provider attributes correlation analysis to their referral frequency and awards. BMC health services research, 16(1), 1-13. https://doi.org/10.1186/s12913-016-1338-1

Zhou, P., Ang, B. W., & Poh, K. L. (2008). A survey of data envelopment analysis in energy and environmental studies. European journal of operational research, 189(1), 1-18. https://doi.org/10.1016/j.ejor.2007.04.042

Información adicional

Clasificación JEL: C4; C5; C6