POLSC401: Ethics and Public Policy

5.1. Benefits and Taxes

Discussions of how policy affects income inequality usually begin with the government benefits and taxes that modify market outcomes. In all rich democracies, the sum total of what governments distribute, often on a means-tested basis, in benefits and extract in taxes ameliorates income inequality. The reduction in inequality from these tools is not especially pronounced in the United States. Clearly, since the late 1970s, the redistributive effects of benefits and taxes have not kept pace with growing inequality in market outcomes before benefits and taxes.

In 1988, Ronald Reagan famously observed, “We fought a war against poverty, and poverty won.” While Reagan may have the better quotation, in fact, government benefits do reduce poverty,^[23] and decreases in benefits increase poverty. The 1996 welfare reform replaced the major income support, Aid to Families with Dependent Children (AFDC), with Temporary Assistance to Needy Families (TANF), which is time-limited and, on average, provides a less generous level of support. While the welfare rolls were already declining at the time that welfare reform was passed, the drop in coverage has been marked: in 2014, the TANF caseload was only 27 percent of what it was at the height of the old welfare program in 1994 (Edin and Shaefer 2015, p. 7).

That a much smaller share of poor families receive TANF benefits and that inflation-adjusted TANF benefits are, on average, lower than at the time of welfare reform is at least part of the story of the increase of the proportion of families with children that are deeply poor. The decline of TANF has reduced government benefits that redistribute income, but that is not the end of the story.

The main income support for the poor is now the Earned Income Tax Credit (EITC), a tax benefit that is tied to work. The EITC ameliorates income inequality, but the link to employment means that it only helps those who are able to find jobs. As a result, it has limited countercyclical impact in economic downturns (Bitler and Hoynes 2016).

Two other programs partly fill the gap. During the Great Recession, SNAP – the Supplemental Nutrition Assistance Program, the heir to what was known as “food stamps” – and Unemployment Insurance provided compensatory assistance with the result that, according to a Congressional Budget Office Report, the reduction in income inequality from government benefits was somewhat higher in 2011 than it had been in 1979.^[24] With economic recovery and the expiration of the temporary boost to SNAP benefits, SNAP spending has declined and with it the temporary boost in the redistributive impact of government benefits (Rosenbaum and Keith-Jennings 2016).

The American tax system is complicated. While it is broadly progressive, in certain respects, it is friendlier to those with higher incomes. Most people’s federal taxes are fairly straightforward and are withheld from their paychecks.^[25] In contrast, for those with complex investments and financial dealings, the federal tax code is riddled with tax-reducing deductions and loopholes inserted by lobbyists and exploited by lawyers and accountants. Because such provisions are so byzantine, they are open to legitimate differences of interpretation as well as flagrant dishonesty. Significantly, in the late 1990s, Congress passed legislation crippling the tax enforcement capacity of the IRS and has subsequently eroded its funding. Nevertheless, after federal taxes, income inequality is somewhat less pronounced than before the federal government takes its bite.

Unlike federal taxes, state and local taxes, which vary substantially from state to state, are not progressive overall. With each rung up the economic ladder, the proportion of income paid in state and local taxes diminishes. In 2015, the share of income paid by the lowest 20 percent of households, those with incomes below $19,000, was 10.9 percent; by the middle 20 percent, those with incomes between $35,000 and $56,000, 9.4 percent; and by the top one percent, who had incomes over $471,000, only 5.4 percent (Institute on Taxation and Economic Policy 2015).

After decades of tinkering with the federal tax code, federal taxes now do less to ameliorate income inequality than they did in 1979.^[26] A number of the alterations to federal taxes have had a regressive impact.^[27] Compared to those further down, the affluent realize a higher share of their incomes from unearned income – that is, income like rents and dividends not derived from wages and salary. Therefore, the successive reductions of the capital gains tax to 15 percent have, on average, reduced the tax bills of those at the top of the income ladder. Similarly, changes to the federal estate tax have reduced estate taxes for the very wealthy.^[28]

Perhaps most notably, income taxes on high salaries have been reduced substantially. Beginning in the 1980s, most affluent countries lowered marginal income tax rates, but the U.S and the U.K., which had relatively higher rates during the 1960s, went further than most.^[29] During the Reagan administration, the marginal federal income tax rate fell from 70 percent to 28 percent. Under Clinton, it rose to 39.6 percent before being lowered to 35 percent under George W. Bush.

In fact, the most important effect of reductions in the marginal rate on earned incomes may be less on who pays what in taxes than on the pay itself. The era of lower marginal income tax rates has also been the era of soaring compensation. When federal taxes gobbled so much of the high pay, there was less incentive to try to extract the last dollar. With lower marginal rates, the payoff for demanding a big raise has skyrocketed.^[30]

5.2. Government Policy and the Shaping of Market Outcomes

Less widely discussed than the way that government benefits and tax policies modify market incomes is how profoundly government policies shape the operations of markets. Beginning with the capacity of governments to enforce the contracts upon which market exchanges rely, capitalist systems are embedded in a myriad of policies that shape their functioning.

Two issues at stake in the 2016 presidential election – immigration and trade agreements, both of which have implications for economic distribution – are shaped fundamentally by policy decisions. They are but two of the many matters germane to economic inequality in which both market operations and government policies have consequences. That is why, according to Joseph Stiglitz, we must understand “the array of laws and policies that lie beneath the surface – the rules that determine the balance of power between public and private, employers and workers, innovation and shared growth and all the other interests that make up the modern economy.” (Stiglitz 2015)

As with immigration and trade, policies with the potential to enhance workers’ paychecks often have politically powerful opponents and inevitably involve tradeoffs among valued outcomes. Furthermore, agreement on a desired result may not be matched by consensus among economists on how to achieve it.

One policy, not always cited in discussions of growing economic inequality, is the use of fiscal and monetary tools to maintain full employment and, thus, promote higher wages. Slack labor markets tend to place workers, especially low-wage workers, at a disadvantage. Since the Reagan years, economic policy has not always made full employment a priority (Baker and Bernstein 2013; Mishel et al. 2014, pp. 26–31).

Another rarely cited partial explanation for growing economic inequality is reduced competition among firms in various economic sectors. When fewer employers are competing for workers, employers gain leverage in setting wages. Concerns about the impact of mergers on competition for employees, and therefore on wages, are notably absent from merger complaints (Council of Economic Advisers Brief 2016). Besides, antitrust enforcement has not been especially vigorous in recent decades.

A policy that is mentioned more frequently is the minimum wage which, by placing a floor under wages, has a mildly equalizing effect on earnings. Unlike many policies ranging from Social Security benefits to the cap on payroll taxes, the minimum wage, established in the Fair Labor Standards Act of 1938, is not indexed to inflation or to the median wage. The value of the minimum wage peaked in real terms in 1968 and has declined 24 percent since then, even though the half century since then has witnessed substantial economic growth (Elwell 2014, Table 1).

It is also not especially generous in comparison with other affluent democracies. In a group of 13 affluent countries, only in Spain and Japan is the minimum wage a smaller percentage of median earnings than in the United States (Atkinson 2015, p. 149). Therefore, what the government has not done – or, at least, what it has not done very often since the Reagan Administration – raise the minimum wage, has contributed to income inequality.

Overtime pay presents a parallel case. Like the minimum wage, guaranteed overtime pay, mandated in the Fair Labor Standards Act, is not indexed. Because the threshold for overtime eligibility is not automatically adjusted with inflation, the share of salaried workers who qualify for overtime pay had sunk to 11 percent in 2014 – from 65 percent in 1975.^[31]

In an era when workers have, on average, reduced leverage in bargaining with employers, employers have adopted a number of practices to keep labor costs down. Among them are hiring outside contractors to do work once undertaken by regular employees who qualified for such fringe benefits as health insurance; requiring employees with complaints about pay or employment practices to submit to binding arbitration rather than to sue in court; enjoining employees from discussing matters of pay with one another; failing to pay mandated minimum wage or overtime; and requiring new employees – not just engineers in Silicon Valley who might have access to trade secrets but such low-wage employees as fast-food workers or camp counselors – to sign non-compete agreements limiting their freedom to seek better-paid jobs.^[32]

All of these practices could be modified by government policy. In some cases, policy makers have chosen to eschew any policy remedy or have been met with successful opposition by affected business interests. In others, the policy tools are in place but inadequately enforced. For example, when it comes to paying below minimum wage or depriving employees of overtime, the number of federal inspectors was cut by nearly a third between 1980 and 2007. Even with the reduced capacity to enforce wage and overtime guarantees, more than $1 billion of stolen wages were recovered in 2012, a figure that is thought to be a small fraction of the national total (Stiglitz 2015, p. 47).

The growth of the financial sector and the explosion of its profitability and the compensation of its higher-ups along with consumer losses during the Great Recession clearly exacerbated income inequality. Nevertheless, it is difficult to assess the contribution to that outcome of government action and inaction with respect to investments, borrowing, and other financial transactions.^[33] Political conflict about government regulation of finance tends to involve lobbying by organized interests – for example, credit card companies, banks, and other financial institutions. As is so often the case, policy impact is buried in the details, details that have been scrutinized and shaped by interests with insider status, policy expertise, and deep pockets.

A number of government actions in the post-2000 era seem to have been particularly friendly to business. For example, the 2005 bankruptcy reform that made it harder for consumers – and nearly impossible for indebted students^[34] – to discharge debts by declaring bankruptcy is surely more advantageous to credit card companies, mortgage lenders, car lenders, and for-profit universities than to indebted consumers. Similar arguments are made about policy developments with regard to the increased protection of intellectual property and the relaxation of antitrust enforcement (Reich 2015).

Much of the story of the financial sector in recent decades, however, involves the absence of regulation. The 1999 repeal of the 1933 Glass-Steagall Act, which had separated investment banking from commercial banking, involved explicit deregulation. However, the absence of regulation can also reflect successful industry opposition, as in the case of the complex financial instruments known as derivatives, as well as the failure of enforcing agencies to regulate vigorously – whether out of ideological conviction, coziness with the industry, or insufficient budgets.

Combined with technological developments that transformed how financial transactions take place, this unregulated environment incubated new ways of doing financial business. Speculation in arcane and complex financial instruments and other forms of financial risk-taking, predatory lending and, sometimes, actual fraud ended up jeopardizing the solvency of financial institutions and leaving many consumers indebted or foreclosed.

When the collapse of major financial institutions was imminent, the government came through with a bailout that, while eventually repaid in full, cost more than the government spent on the unemployed who lost jobs during the ensuing recession and helped bankers retain their bonuses while leaving behind those who had lost homes to foreclosure. In the aftermath, Congress legislated new financial regulations in the 2010 Dodd-Frank Wall Street Reform and Consumer Protection Act. Although passed with a party-line vote, the bill involved substantial compromise. Both the bill itself and its implementation, which has not been especially forceful, have been controversial. With the recent transition to the Trump administration, the fate of Dodd-Frank remains in doubt.

5.3. Declining Unions and Growing Economic Inequality

Any account of the growth of economic inequality must include a discussion of the decline of labor union membership and power, a development with consequences for both economic and political inequality. Unions operate in several ways to boost workers’ power and enhance their earnings. Union members are more likely to be politically active even beyond what we would expect on the basis of their education and incomes. Furthermore, labor unions represent their members’ economic interests – and the economic interests of ordinary citizens more generally – in politics. Unions also provide workers a collective voice in the workplace.

The share of the workforce that is unionized actually peaked in the 1950s, but the past three decades have witnessed striking attrition in the proportion of workers who are union members and the slightly higher proportion who are covered by union contracts. It is notable that, even as the size of the workforce expanded substantially, the absolute number of union members declined by nearly five million over this period. In 1977, 26.5 percent of all wage and salary workers were members of unions; by 2014, the figure had dropped to 11.1 percent.^[35]

America’s circumstances are not unique and over the last generation, the erosion in the proportion of the workforce that is unionized has been common across industrial democracies. In a group of twenty advanced democracies, the proportion of unionized workers diminished between 1979 and the late 1990s in fourteen of them.^[36] Still, even if the United States is hardly alone in the drop in union density, the United States has, in comparative terms, very low levels of union membership.

The decrease in union ranks has occurred entirely in the private sector. The proportion of private-sector workers who were union members decreased steadily, from 21.7 percent in 1977 to 6.6 percent in 2014. In contrast, the share of public-sector workers who are union members fluctuated within a very narrow range and ended the period at a slightly higher level, 35.7 percent, than at the beginning. While the share of the workforce employed in the public sector fell from 18.7 percent to 15.2 percent over this period, the share of union members who are public-sector employees rose sharply, from 31.4 percent to 54.4 percent.

There is no question that the decline of union membership has had an impact on the growth of income inequality, although there is disagreement about how much. The union wage premium – that is, the increment to wages and benefits accruing solely from union membership with other relevant factors taken into account – has diminished since the 1980s. Still, union membership clearly boosts compensation – especially for private-sector employees, for men, for blue-collar workers, and for workers with no post-secondary education.^[37] In fact, one study estimates that union decline explains between a fifth and a third of the growth in wage inequality among men (Western and Rosenfeld 2011).

At the same time that the weakening of union economic power has had consequences for the size of workers’ paychecks and the conditions under which they work, union decline has diminished their political capacity to support policies that protect the economic interests of ordinary workers and to oppose policies that benefit the privileged. An indicator of the political weakness of organized labor is the finding in a recent study that when asked to name their principal antagonists on the issues on which they were currently working, not one of the corporate lobbyists mentioned a union (Drutman 2015, pp. 78–79). Thus, union decline has operated through both the workplace and politics to enhance income inequality.

What explains the steep decline in the share of American workers who are union members? A number of factors account for this trend.^[38] One is structural changes in the American economy, in particular the decrease in manufacturing employment. Another is diminished support for unionization among workers. Unions themselves made miscalculations. There is evidence that the leadership of the American Federation of Labor and Congress of Industrial Organizations (AFL-CIO), especially George Meany, did not devote sufficient attention or resources to organizing. Furthermore, employers have been increasingly aggressive. Aided by consultants who specialize in “union prevention,” businesses have become substantially more hostile to union-organizing drives in both tone and tactics.

Finally, the weakening of labor has undoubtedly had a political component.^[39] A significant blow to organized labor unions was the passage in 1947 of the employer-friendly Taft-Hartley Act, which proscribed a number of labor practices and permitted states to pass “right-to-work” laws outlawing the union shop. Although attempts have been made to alter or repeal it, Taft-Hartley remains in place today. In fact, organized labor has not succeeded in realizing any of its major legislative goals in decades. During both the Carter and Obama administrations, Congresses controlled by Democrats handed legislative defeats to organized labor.^[40]

Political developments in the early years of the Reagan administration are critical. In a turning point in labor history, during the summer of 1981, Reagan dismissed striking air traffic controllers and replaced them with nonunion employees, after which employers have felt free to replace striking workers. In addition, Reagan was able to name appointees to the National Labor Relations Board (NLRB) who were unfriendly to labor. The result was policy changes that have facilitated management’s capacity to act aggressively against unions.^[41]

The NLRB was able to weaken worker protections under the National Labor Relations Act by overturning worker-friendly precedents, many of them long standing, through a series of decisions in carefully selected cases. At the same time, whether by accident or design, the number of decisions in cases of unfair labor practices dwindled, and the backlog of unresolved cases expanded to the largest number in history (Gross 1995, p. 253). More recently, at the state level, bolstered by friendly state legislatures, Republican governors in a number of states have dealt with revenue shortfalls by cutting the medical benefits and pensions of unionized public employees and, more fundamentally, targeting their collective bargaining rights.

6. Conclusions

Economic deserts are not only unequally distributed but are more unequally distributed now than at any time in several generations and are more unequal in the United States than in most developed democracies. Those at the top have garnered most of the income gains while incomes in the middle and lower ranges have stagnated, and the number of desperately poor has risen.

What we have seen undermines several clichés about economic life in America. For one, it is often argued that a rising economic tide lifts all boats. However, growing prosperity and productivity in recent decades have lifted the yachts but left the dinghies still grounded. Moreover, with the stagnation in middle- and lower-class income and earnings, no longer does American affluence imply that low-income workers are better off in absolute terms than they are elsewhere.

Finally, in spite of rhetoric about America as the land of opportunity, well-educated, affluent parents are ordinarily able to pass their high status along to their children, and the rates of upward mobility in the United States are actually lower than in most affluent democracies. Besides, the possibilities for those of modest origins to become successful have surely not increased in this era of increasing economic inequality. Indeed, they may have diminished.

It is difficult to sort out the causes of economic inequality, but there is agreement that technological change, international trade, and domestic outsourcing are significant factors. Although the role of politics is more controversial, an environment of government policy shapes the way the markets function, and government policies influence the distribution of income.

In addition, government inaction has exacerbated income inequality. Public officials have had policy instruments at their disposal – for example, raising the minimum wage regularly or establishing prudent standards for mortgage loans – that might have decelerated the rate of growth of economic inequality. Thus, the sum total of government action and inaction has been insufficient to keep up with growing income inequality in the United States and has had less impact on the increase than in other affluent democracies.

Another factor that has added to economic inequality is the decline of unions and the resulting reduction in the economic and political power of workers. Even though private-sector union membership is a fraction of what it once was, unionized workers still command better wages and benefits than otherwise similar non-union workers. The weakness of unions also has implications for the inequalities of political voice. Among organized interests, unions are the most important advocates for the economic needs and concerns of ordinary workers. Labor union membership has a mobilizing effect for individuals: compared to their non-union counterparts, union members are more likely to vote or to engage in other political activity.

We are concluding this essay in the same week as the announcement of the broad outlines of a tax plan that, in one assessment, “shifts trillions to the wealthiest.” (Davis and Cohen 2017). Should this tax plan be enacted in roughly its proposed form, it would hardly ameliorate the current levels of economic inequality and would serve as additional evidence that politics matters for economic outcomes.

Notes:

1. Walton Family Net Worth is a Case Study Why Growing Wealth Concentration Isn’t Just an Academic Worry. Economic Policy Institute Working Economics Blog, posted October 3 (Bivens 2014).

2. The minimum wage for 1978 is found at U.S. Department of Labor, Wage and Hour Division. 1938–2009. (Wage and Hour Division 2015); cost of living adjustment is taken from U. S. Department of Labor, Bureau of Labor Statistics (2015), CPI Inflation Calculator; the rate of growth of CEO pay (including the value of stock options exercised in a given year plus salary, bonuses, restricted stock grants, and long-term incentive payouts) for chief executives of the top 350 U.S. firms is taken from Lawrence Mishel and Alyssa Davis. (Economic Policy Institute 2015).

3. Data are for the 34 members of the Organization for Economic Cooperation and Development (OECD). The poverty rate is the proportion of the population whose incomes are below half the median for the population as a whole. (OECD 2014, pp. 66–67)

4 See, for example, (Brewer 2007; Bartels 2016; Stonecash 2010, chp. 7).

5. For extensive additional bibliography and discussion of technical matters, see (Schlozman et al. 2012, chp. 3). See, also, (Piketty and Saez 2003, pp. 1–39; Burtless and Jencks 2003, chp. 3; Mishel et al. 2012; Piketty 2014, esp. Part III; Atkinson 2015, chp. 1).

6. There is controversy among economists about the long-used official definition of poverty. An alternative measure, which takes account of in-kind government benefits, shows lower rates of poverty among children and higher rates among adults, especially the elderly. See (Bridges and Gesumaria 2015, pp. 55–81; Short 2015, pp. 60–254).

7. On desperate poverty, see (Edin and Shaefer 2015).

8. (Economic Policy Institute 2015). The figure is for chief executives of the top 350 U.S. firms and includes the value of stock options exercised in a given year plus salary, bonuses, restricted stock grants, and long-term incentive payouts.

9. See (Bebchuk and Fried 2004)

10. Cited in (Reich 2012, p. 11)

11. Cited in (Stiglitz 2013, p. 257).

12. On the erosion of the private welfare state, see (Katz 2001, chps. 6–8; Hacker 2006).

13. Figure taken from Employment Benefit Research Institute (2015, Table 5.1a), EBRI Databook on Employee Benefits.

14. Discussion in this paragraph is taken from (Mishel et al. 2012, pp. 376, 385–95).

15. This section draws on arguments and data in (Burtless and Jencks 2003; Smeeding 2005; Piketty 2014, especially chps. 8–9; Atkinson 2015, chp. 2). Making cross-national comparisons with regard to these issues poses technical dilemmas. See these sources as well the discussions and citations in (Schlozman et al. 2012, pp. 76–79).

16. (Smeeding 2005, pp. 971–73); See, also (Krueger 2012, Figure 10); and (Gornick and Milanovic 2015).

17. In a vast literature, see, for example, (Blau and Duncan 1967; Hauser and Featherman 1977; Hout 1988, pp. 1358–400; Ganzeboom et al. 1991, p. 284; Burtless and Jencks 2003); the essays in (Bowles et al. 2005); and the essays in (Isaacs et al. 2008).

18. On the class gaps in child well-being, see (Putnam 2015). With respect to the educational system, see the essays and references in (Duncan and Murnane 2011a), in particular, (Duncan and Murnane, “Introduction,” chp. 1; Reardon, Sean F. “The Widening Academic Achievement Gap between the Rich and the Poor: New Evidence and Possible Explanations,” chap. 5; and Bailey, Martha J., and Susan M. Dynarski, “Inequality in Postsecondary Education,” chp. 6.)

19. An exception to the pattern of growing gaps between rich and poor children is diminution in the disparity in health between rich and poor for children and young adults. See (Sanger-Katz 2016)

20. (Corak 2013, Figure 1). See also, (Krueger 2012).

21. Discussion of such factors is contained in (Dadush et al. 2012, chp. 4). See also, (Atkinson 2015, chp. 3).

22. Testimony before the Massachusetts legislature, cited without additional bibliographic information in (Kens 1998, p. 19). On the extent to which executive pay reflects forces other than the operations of markets, see (Bebchuk and Fried 2004, Parts I and II).

23. For evidence and citations supporting the contention that government benefits reduce poverty, see (Ziliak 2015, pp. 34–36).

24. (Congressional Budget Office 2014, pp. 25–27). A parallel CBO analysis undertaken three years earlier found the opposite, a decrease in the redistributive impact of government benefits, for the period between 1979 and 2007. See (Congressional Budget Office 2011, Xii).

25. Material in the paragraph is taken from (Johnston 2007, pp. 168–73).

26. (Congressional Budget Office 2014, Figure 15). See also, (Fieldhouse 2013).

27. Information in this paragraph is taken from (Stiglitz 2013, pp. 89–92); and (Formisano 2015, pp. 77–80).

28. For information about the estate tax, see (Jacobson et al. 2007, volume 2, chp. 1); (Jacobson et al. 2007) and (Huang and DeBot 2015).

29. (Atkinson 2015, pp. 181–82). See also, (Piketty 2014, pp. 499, 508).

30. Among others, this argument is made by Piketty (2014, pp. 508–12), who finds no evidence that the explosion in compensation has been accompanied by enhanced productivity by high earners.

31. (Mishel and Eisenbrey 2015, p. 10). In the final year of the Obama administration, the Department of Labor put out a rule making many additional workers eligible for overtime pay. Immediately after Trump took office, this rule was suspended. See (Opfer 2017).

32. These and other practices are discussed in (Stiglitz 2015) and (Council of Economic Advisers Brief 2016). The specific examples of lower-wage workers who are required to sign non-competes are in (Council of Economic Advisers Brief 2016, p. 8).

33. For conflicting views on the causes of the 2008 financial crisis, see (Financial Crisis Inquiry Commission 2011), including the two dissenting reports. Material in the remainder of this section is taken from (Warren and Tyagi 2003, pp. 126–29, 152–56; Stiglitz 2013, pp. 46, 93, 112–15, 239–45, 252, 310; Reich 2012, pp. 57–58).

34. One provision in the 2005 bankruptcy act made it extremely difficult to discharge private student loans—in contrast to, for example, consumer debt—through bankruptcy. See (Lieber 2010).

35. Data taken from the Union Membership and Coverage Database constructed by Barry Hirsch and David Macpherson. (Hirsch and Macpherson 2015)

36. (Flanagan 2005, p. 35. Table 1). See (Mishel et al. 2009, p. 375). Of the thirteen countries about which they present data, union coverage is lowest in the United States.

37. See (Mishel 2012; Rosenfeld 2014, chps. 2–3).

38. For a more extensive discussion and additional bibliographical sources, see (Schlozman et al. 2012, pp. 87–94); as well as (Freeman 2007, chp. 5). See also (Goldfield 1987; Freeman and Katz 1994; Hacker and Pierson 2010, pp. 56–61; Rosenfeld 2014, chp. 1).

39. On these factors, see (Lichtenstein 2002, chps. 3–4).

40. See (Hacker and Pierson 2010, pp. 127–32, 278–79).

41. On the NLRB under Reagan, see (Levy 1985, pp. 269–390; Moe 1987, pp. 266–71; Gross 1995, chp. 13).