Risk-pooling--necessary but not sufficient?

The search for a ‘‘good’’ health systemis constant. Even in countries wherecomprehensively organized national healthsystems have existed for over a century,the debate on how best to design them toensuretheeffectivefinancing,organizationand provision of care is as lively andcontroversial as ever.During the last quarter of the20th century many countries proposed awide variety of sweeping reforms, someof which were adopted and refined, andothers rejected. That period can be viewedas one in which a ‘‘natural experiment’’was conducted on a global scale. It hasyielded surprisingly little, however, in theway of real, evidence-based understandingof what works well.Much of current research on healthpolicy seeks to discover quantitative rela-tionships that can explain the achievementof health systems in relation to their designand the wider context within which theyoperate. One such relationship which isalready quite well understood concernsthe role played by risk-pooling, or insur-ance (1).Thereisagrowingconsensusthat,other things being equal, systems in whichthedegreeofrisk-poolingisgreaterachievemore.Risk-pooling is beneficial becausehealthcarecostsaregenerallyunpredictableandsometimeshigh.Peoplecannotreliablyforecast when they will fall ill and needto make use of health services. When ithappens, the costs of those services can besignificant. Risk-pooling increases the like-lihood that those who need health carewillbeabletoobtainitinanaffordableandtimely manner. It allows resources to betransferred from the healthy to the sick.From the viewpoint of individuals andhouseholds, contributions during timesof good health can be used to meet healthcare costs in the event of illness. In manycases, pooling also contributes towardsredistributive goals by making those withhigher incomes contribute more in orderto subsidize the poor.Community-basedhealthinsuranceor‘‘micro-insurance’’(2) is one approach torealizing the benefits of risk-pooling incountries without the organizational capa-citytooperateformalschemesonanationallevel. Risk-pooling, on any scale, has itsdownside, however. Clients who areinsured and therefore do not have to paythe full cost of services may be inclinedto over-use those services, while providersmay be happy to let them do so because itenables them to earn more (moral hazard).In addition, those who are more likely torequirecarehaveastrongerincentivetojoina voluntary risk-pooling scheme (adverseselection).Suchproblemscouldbereducedin micro-insurance schemes since theirsmall scale and community focus mightprovide informal safeguards against them.Smallnessinitsturnisamixedblessingthough. As Dror explains in this issue ofthe Bulletin (pp. 672–678), the smaller ascheme is the harder it is for it to remainsolvent. That is a statistical reality whichcannotbeavoided.Hesuggestsreinsuranceas an approach to help micro-insuranceschemes overcome the threat of financialfailure. Just as households can pool theirresourcestohelpreducefluctuationintheirhealth spending, so can insurance schemesthemselves. Throughreinsurance, schemesthat have faced fewer demands on theirfunds than anticipated can indirectlysubsidize those that have faced more.The potential benefits of reinsurancein countries where micro-insurers operateare clear. Unfortunately, however, reinsur-ancealsofacesthetwindifficultiesofmoralhazard and adverse selection. Micro-insur-ance schemes that reinsure may haveweaker incentives to contain the costs ofthebenefitstheypayout;andschemesthatare more likely to face budget blow-outs inany given period are more likely to chooseto reinsure.Reinsurance is attractive because itexpandsthesizeoftheriskpool,andallowsgreater scope for costs to be shared. Insome circumstances, however, it may bepreferablesimplytoestablishschemeswithlargerriskpoolsfromtheoutset.Examplesalready exist of district-based health insur-ance (3), targeting tens of thousands ofpeopleratherthanasmallcommunity.Suchlarger schemes can also benefit fromeconomies of scale in administrationand transaction costs.Alternatively it may be feasible toencourage progressive scaling up of micro-insurance schemes, possibly through aplanned programme of mergers, to formever larger risk pools. Such an approachcould allow infrastructure and other cap-abilities to develop over time, and mayeventuallyleadtotheemergenceofasingleregional (or national) risk pool whichmany would consider to be the ideal.In addition, attempts to employ risk-pooling will fail if they do not have theconfidenceofthepeopletheyseektoserve.Any pooling scheme will be judged by itsresults, and if problems of access oraffordability persist people will quicklybecome cynical. Appropriate arrangementsare needed to oversee the scheme so asto ensure that the benefits of risk-poolingare in fact realized.More fundamentally, those who paycontributions when they are healthy mustknow that, if and when they need care, thepooling scheme will still be there to helpthem. The failure of an insurer can be ascatastrophic for those affected as thecollapseofabank.Countriesmaythusneedto regulate insurers and micro-insurers toensure financial probity in the same waythattheyregulatethebankingindustry.Theimportance of oversight and regulationfor effective risk-pooling should not beoverlooked.Similarly, other initiatives to improvehealth system performance — whetherby increasing the efficiency of serviceproviders, building up human resourcesor enhancing access to traditional healing— will struggle where government is weakin its role as protector of the public’sinterests.Successinthatrole,definedbytheWorld Health Report 2000 — health systems:improving performance as stewardship, is un-doubtedlyakeyfactorindetermininglevelsof health system achievement. In contrastto the concept of pooling risk, however,little evidence currently exists as to whatconstitutes effective stewardship, how tobringitaboutandthemechanismsbywhichit has an impact on achievement. Thereinlies another key challenge in understandingthe complex relationships that link policyand performance.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.