Academic journal article Journal of Risk and Insurance

Mitigating Insurance Fraud: Lump-Sum Awards, Premium Subsidies, and Indemnity Taxes

Academic journal article Journal of Risk and Insurance

Mitigating Insurance Fraud: Lump-Sum Awards, Premium Subsidies, and Indemnity Taxes

Article excerpt


Governments have many tax tools at their disposal to redistribute wealth from one segment of the population to another, such as income, sales, and lump-sum taxes (for example, setup fees for corporations). This article studies the interaction between three types of taxes in an economy plagued with insurance fraud. The author presents a case where an excise tax on insurance benefits is the preferred form of taxation because it reduces insurance fraud in the economy. The optimal tax combination is to use all the proceeds from the indemnity tax to subsidize premiums and to use a lumpsum tax to fill the government's coffers. This tax scheme allows a government to reach the first best allocation in the limit, even in the presence of market imperfections (ex post moral hazard). When the government's cost of administering the tax scheme increases in the indemnity tax rate (a Laffercurve argument), it remains optimal to redistribute all proceeds from the indemnity tax as premium subsidies; the first best is no longer attainable in this case, however.


One reason that modem governments are unwilling to use lump-sum taxes (also known as poll taxes) may be that to do so corresponds to political suicide. This is true even if a lump-sum tax is generally perceived as the most efficient form of taxation. Although lump-sum taxes are efficient when perfect information is available, this may no longer be the case when information is not perfect. The general argument against a head tax is that it provides no insurance in case of an unlucky draw. However, a wage tax, as in Eaton and Rosen (1980), or a corporate tax, as in Peck (1989), provides some insurance in the sense that agents who have a lucky draw (higher wage or profit) end up paying more in taxes.

The model presented here examines the incidence of different taxes on the welfare of agents when markets are imperfect. The approach the author will take is one where a worker may suffer an injury that prevents him or her from working. This introduces some income uncertainty for the worker. Although the worker faces uncertainty regarding the future the author allows an insurance market--namely a workers compensation or disability insurance market--to exist to insure against that uncertainty. This differs from Eaton and Rosen (1980) and Peck (1988, 1999), where no insurance market exists.

The market imperfection the author introduces is that the existence of the disability is known only to the worker and not known for certain by the insurance company unless it conducts a costly audit of the disability insurance claim. This means that the worker may have the incentive to misreport the true state of the world to extract more money from the insurer. This approach, known as the costly state verification approach, was pioneered by Townsend (1979). Reinganum and Wilde (1985), Mookherjee and Png (1989), Picard (1996), and Bond and Crocker (1997) also use this approach.

In such a setting, the traditional approach has been to say that the insurer can commit to an auditing strategy such that it is in the worker's best interest to always tell the truth. Unfortunately, it may not be credible for the insurer to commit to such an auditing strategy, as argued by Graetz, Reinganum, and Wilde (1986), Picard (1996), Khalil (1997), and Boyer (1998). Suppose the insurer announces an auditing strategy such that workers have nothing to gain by falsely reporting the state of the world. If the worker believes in such an audit strategy, then he or she will always tell the truth. The insurer, on hearing the worker's report, has no reason to audit since audits are costly and will only reveal that the worker has told the truth. By not auditing, the insurer saves the cost of auditing, which means that commitment may not be credible. Consequently, workers, anticipating the insurer's unwillingness to audit, will want to lie. Therefore the principal-agent problem between the insurer and the worker is not solved. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.