One of the biggest problems we are facing in reducing poverty is how to reduce in-work poverty. Structural trends in the labor market make in-work poverty a real problem. They range from the rise of low-wage work, deindustrialization (and the loss of good-paying manufacturing jobs), the rise of service-sector work with insufficient stability (low wages, low hours, unpredictable schedules etc.), to the digitalization of work (Uber, Mechanical Turk, Task Rabbit and others).
Policy discussions that aim at reducing in-work poverty essentially have two solutions in mind. Each of these solutions really have their advantages and disadvantages. One is to raise the minimum wage and the other is to increase the negative income tax, which is called Earned Income Tax Credit (EITC) in the US. I argue that we have to consider other options to alleviate poverty as well, most importantly the universal basic income. But for now, let us reflect on each of these two solutions.
The minimum wage is the minimum amount of wage per hour that can legally be paid to a dependent employee by his employer. That is currently 7.25 dollars an hour in the US, though some jurisdictions have a higher (but not lower) minimum wage.
The rationale behind raising the minimum wage is that workers get paid so little because capitalist bosses can take advantage of the weak bargaining power of their workers (take say Walmart, where unionization drives had been largely unsuccessful) and pay their workers only a small fraction of what they produce. If there were no minimum wage at all, there would be workers that would get paid only 3 or 4 dollars an hour, and workers that are desperate for any kind of employment and not knowing any better places to turn to will accept that low wage. In economic theory, a minimum wage is justifiable under the premise of a monopsonistic labor market, which implies one or few employers and many workers. This imbalance tends to favor employers, who can threaten workers with easy replacement if they ask for a higher wage. In this case, the employer is using his political power to hold down his workers’ wages, and a higher minimum wage merely transfers the rent that otherwise accrues to the employer back to the workers.
Critics of the minimum wage claim that it is an undue intervention in the labor market, which could hurt low-wage workers, because by artificially raising the wages, the number of workers employed is being reduced. The effect is to eliminate low wages, but also to eliminate many low-wage workers from the labor market altogether or push them into the shadow economy. I would counter that this equilibrium condition assumes that there are no excess rents, which firms collect. But it is true that in a minimum wage that is raised too high (e.g. more than 50% of the local median wage, see Dube n.d.), adverse effects on employment will offset other advantages of a hike (like reduced turnover, higher productivity and less firm rents). A July 2016 study of the Seattle minimum wage (Seattle Minimum Wage Study Team 2016) finds ambiguous effects of a hike to 15 dollars an hour, which has been carried out in several steps. The number of business closures has not increased, the employment rate of low-wage workers was reduced slightly, the overall paycheck for low-wage workers increased but they faced small reductions in hours worked.
This evidence seems to propel the view of the critics of the minimum wage, who think that fixing wages in the labor market above the natural rate will result in less employment and fewer hours of work. Starbucks, for instance, promised to raise the wages for its workers, but coupled that with the use of scheduling software, which cut the number of hours for their workers, so they ended up with less pay even as their hourly wage increased (Durden 2016). On the other hand, Starbucks might be a poor example to pick, because they belong to the type of large firm, which collects excess rents from their workers, and could probably easily afford to pay their workers more money.
Market liberals are skeptical about the minimum wage, but they instead tend to be very favorable toward the negative income tax (e.g. Milton Friedman, see Allen 2002). In the negative income tax, low-income earners would not pay any tax, but would receive tax credits. To quote the example in Allen
If, for example, the threshold for positive tax liability for a family of four was, say, $10,000, a family with only $8,000 of annual income would, given a negative tax rate of 25 percent, receive a check from the Treasury worth $500 (25 percent of the $2,000 difference between its $8,000 income and the $10,000 threshold). A family with zero income would receive $2,500.
Since the Reagan administration, the most prevalent form of the negative income tax has been the Earned Income Tax Credit (EITC), which provides most tax benefits to low-wage earners with dependent children (those below the age of 24). The program has been sold as a way to encourage work among able-bodied individuals without distorting the labor market. Let us take a simple example to illustrate this point: there is a low-wage worker, who produces 15 dollars an hour in value to his employer. The employer pays him the minimum wage of 7.25 dollars an hour. If the federal government hiked the minimum wage to 15 dollars an hour, the net profit would shrink to zero, and the worker would have to be laid off. Turning to EITC: If that worker is the primary earner in the family, then having an EITC scheme, which would subsidize his meager earnings by 5,000 dollars annually, would go a long way toward paying part of the rent and other living expenses he faces without changing the nominal wage. There is no need to lay that worker off and his poverty is alleviated. The advantage of the EITC is clearly that there is no assumption about the worker productivity that needs to be made in order to reduce poverty (high productivity would presumably result in a wage high enough to prevent poverty).
But there are problems with a negative income tax, because in the ideal world we imagine that employers don’t react to the creation of a negative income tax, i.e. they don’t change the wages of their workers knowing that the scheme exists. But we know that they do react to negative income taxes by paying their workers below subsistence wages even if they could afford to pay them more. Karl Polanyi (1944: 81-89) had been a major critic of the British Speenhamland law, which was a form of poor relief similar to the negative income tax. The poor workers would be subsidized by municipalities so they can at least afford bread. Once the law was implemented, the employers milked the law by deliberately paying below subsistence wages, knowing that their workers would receive a top-up from the municipality. The growing burden of low wages implied that municipalities would not be able to fund generous top-ups, and they were eventually forced to reduce the top-ups, which created misery among the workers. The law was later replaced by a no less draconian poor law, which created workhouses to treat the “undeserving” poor that were otherwise “unwilling” to work.
If we think about what Walmart is doing in today’s economy, it can’t be much different from what employers did in 19th century Britain. Walmart is taking a classical “low-road strategy” (Kalleberg 2011) of paying their workers very low wages, even though they could easily afford to pay them more given their huge share buybacks. Traub (2015) writes,
If Walmart redirected the $10 billion per year it has authorized for buybacks toward investment in human capital, it could provide its 825,000 lowest-paid U.S. employees a raise of as much as an additional $7.67 per hour without raising consumer prices by a penny. On top of the $10 an hour Walmart has already committed to, this would more than pay for the $15 an hour Walmart workers are calling for.
Instead, Walmart continues to pay low wages to their workers, knowing that they get Food Stamps, Medicaid and EITC from the taxpayers. A report by Jacobs et al. (2015) finds that the US taxpayers are subsidizing Walmart, McDonalds and other low-wage employers to the tune of 152.8 billion dollars per year. Only a part of that goes to EITC, which is the focus of this post, but it is clear that employers, who have the ability to milk the government and middle class families, will do so without impunity.
I have now shown that neither the minimum wage nor the negative income tax are ideal solutions to alleviate poverty at the minimum social cost among working people. The minimum wage is a direct intervention in the labor market and has the capacity to reduce the rent going to employers, but has the risk of making labor too expensive to hire, which is a genuine fear in largely unproductive service sectors.
The negative income tax can bypass the direct labor market intervention and this risk, but will encourage employers to pay their workers too little. Here the risk is that to eliminate poverty either the tax benefits have to be so large that the government program becomes unsustainably expensive (also politically difficult to sustain because middle class people don’t benefit from the scheme and will be likely to attack it, which is the logic of the paradox of redistribution, see Korpi and Palme 1998), or the government has to sacrifice their poverty reduction targets. The employer incentive to underpay workers results in a lack of incentive to invest in high-productivity jobs, which would entrench the low pay trap of these workers. In that sense, even the negative income tax does not escape the trouble of intervention in the labor market.
My critique of the minimum wage and the negative income tax by no means implies that I would want to end these two programs. In the absence of better policy alternatives, it would be cruel to downgrade these two programs regardless of what the efficiency effect is. Any government policy to improve the lot of the lower class deserves support if we care about reducing poverty. But we have to look at pro-poor policies as a package consisting of a combination of different policies. I conclude with four ideas: productivity investment, public-sector jobs, unionization and the universal basic income.
We need to invest in the creation of higher productivity jobs out of which higher wages can be easily paid. Again, we assume that there is no mechanism by which employers can prevent the payment of higher wages despite sufficient productivity (e.g. Walmart). There is naturally a risk with high-productivity sectors, which is that they don’t offer that much employment opportunities. There is only so many cars that a country needs, and yet this is a sector where productivity enhancement has been an important part of the operation, resulting in fewer total jobs in the industry. That is why we have long employment lines in the fast food joints and school janitorial positions.
Part of the credential narrative (“get an education to get a better job”) is precisely the expectation that workers can easily shift from low to high productivity jobs. We know that our society is now actually suffering from overcredentialing, i.e. too many highly educated people but not enough jobs for them. The labor market only supports so many accountants and lawyers. Culturally, it is also questionable to demand of so many people to be “smarter” and be more “adaptable” to the needs of the more “complex” labor market. The Atlantic has called this discourse rightfully “the war on stupid people” (Freedman 2016).
Part of the reason why China was able to lift out so many people out of poverty is not because they have accelerated education credentialing (which did not pick up until the late-1990s, long after the economic reform of the 1970s), but because they became the factory of the world, offering so many landless peasants low-skilled factory jobs. If our solution to the in-work poverty problem is work-centric, then the expansion of low-skilled (but probably subsidized) work is what we would need. Expanding high-productivity sectors with high levels of employment is likely only feasible if we can suddenly find new markets to penetrate.
Another work-centric solution would be to increase public-sector employment. The logic would be as follows: first, the government would mainly create jobs with a living wage. Second, because the public now competes for workers with the private sector, the private sector will now also have to raise wages to stay competitive. Again, I expect that the rent function in the private sector decrease. The complaint about public sector jobs is not so much the increase in debt, because the government could increase taxes on the rich without taking out that much debt. Instead of wasting tax dollars on subsidizing low-wage workers, the state could use some of that money for direct job creation instead. The real complaint is of a rather political nature as Michal Kalecki (1943) remarked: basically, capitalists hate the heightened political power accruing to the working class as public-sector job creation swells, undermining the collective class power of the capitalists. In Kalecki’s own words,
[T]he maintenance of full employment would cause social and political changes which would give a new impetus to the opposition of the business leaders. Indeed, under a regime of permanent full employment, the ‘sack’ would cease to play its role as a ‘disciplinary measure. The social position of the boss would be undermined, and the self-assurance and class-consciousness of the working class would grow. Strikes for wage increases and improvements in conditions of work would create political tension. It is true that profits would be higher under a regime of full employment than they are on the average under laissez-faire [emphasis added], and even the rise in wage rates resulting from the stronger bargaining power of the workers is less likely to reduce profits than to increase prices, and thus adversely affects only the rentier interests. But ‘discipline in the factories’ and ‘political stability’ are more appreciated than profits by business leaders. Their class instinct tells them that lasting full employment is unsound from their point of view, and that unemployment is an integral part of the ‘normal’ capitalist system.
Even if capitalists would benefit from full employment following public job creation, they don’t like the political aspect of strengthening the interests of the workers. I would say that public-sector job creation has to return as a tool to reduce poverty and provide opportunities to more workers. The capitalists will get used to their somewhat diminished stature, and, in any case, some capitalist will make investments if the profit is large enough.
Strengthening the political hand of the workers is precisely what is needed in today’s economy, where most young people don’t even know what a labor union is or does. The rising political consciousness of the working class would be associated with the creation of new labor unions. Unions have been in precipitous decline since the end of World War II. I don’t think it has to do with a substantial change in American values (though values certainly shift in the absence of unions). Their decline is associated with the shift from industrial to service employment and the concomitant difficulty of organizing the latter into unions. Nascent efforts like those in Walmart are easily crushed by ruthless human resource practices, like shutting down an entire store trying to unionize. Unlike port workers (the Longshoremen are notorious for their strikes and high pay), most service-sector jobs don’t have a bottleneck position, i.e. their strike does not shut down commerce, so the bosses can ruthlessly go against workers trying to unionize. Union weakness was perfectly exploited by employers eager to increase their profit share. The shareholder value maximization and escalating CEO pay would be inconceivable in a world where unions still had a substantial say in corporate governance (Lazonick 2011).
But these aforementioned solutions are still very much focused on work and the labor market, while technological innovation that is labor-saving has the contrary effect of making it increasingly difficult to find employment opportunities for the masses. Needless to say, structurally greater unemployment (or a decline in the labor force participation rate expressing the same) puts a downward pressure on wages, as more workers compete for the few scarce jobs. Some people might say that Frey and Osborne’s (2013) prediction of eliminating half of all jobs because of automation is exaggerated.
But even if their claim is exaggerated: just eliminating 10 or 20% of all employment opportunities would have substantial knock-on effects in the economy, which will be too slow in finding the people new employment. It would also be questionable to arbitrarily find them work just so they can work. To some extent, our crazily GDP- and employment-growth focused economy has resulted in the proliferation of goods and services, which do not necessarily enhance collective well-being (e.g. every year a slightly different smartphone). We are really creating a lot of make-work rather than useful jobs, and our creativity would have to take another turn in equipping people with new make-work as much of the useful employment is taken up by robots.
We have to sever the tie between employment and income, and only the universal basic income can accomplish that goal. It would be a substantial economic investment, and I have argued elsewhere that we could afford it and it would not have adverse effects on incentives for work, but would liberate people to choose their own employment, whether it is paid or not (Liu 2015a, 2015b, 2016). The real innovation in social policy would not be to tinker with the existing regimes of poverty reduction, though it beats an alternative of inaction, but to promote the universal basic income.