I don't think there is a numerical upper limit for the minimum wage, because the minimum wage should be relative to the cost of living in the particular jurisdiction at the particular time. Some say a higher minimum wage can drive up inflation, but it is more likely that those at the top salary levels drive up inflation. Minimum wage workers tend to spend money locally and on practical needs. 

So are you saying there is absolutely no upper limit, even if you account for the local cost of living?  There must be some upper-limit, even in areas with a high cost of living, right?  

I think other negative effects of a high minimum wage include companies reducing the number of hours worked and hiring less total workers and hiring more part-time workers to avoid paying benefits.

It can also slow down economic growth and discourage new small businesses.

It can also force companies to replace unskilled workers with automation and technology.

Do you agree with these negative effects of a high Minimum wage?  Why or why not?

The negative effect on raising the standard minimum wage on employment is more evident in some socio-professional categories, such as young workers, for example. Also, an increase in the minimum wage can reduce the salaries of poorly prepared workers due to their low productivity, leaving them outside the minimum wage coverage area and increasing the labor supply in the informal sector, and thus lowering the salary in this section.

Thanks for your answer, Daniela.  You make some good points. When the Minimum wage is reltively high, do you think companies might become very selective in who they hire?  

I never thought of that effect, but I think its a very good point.

That might lock out the very unskilled and inexperienced workers.

Yes! I think that in this case, when minimum wage is relatively high, the following situations are possible for workers who produce below the minimum wage: 1 will be fired; 2. will be replaced by cheaper production factors - capital; 3. the employer will no longer engage in employment and will change the organization of work and production; 4. Provide employees with low productivity to the opportunity to form and become more productive with a higher level of skills and abilities.

There is no upper limit of compensation for CEOs and highly paid workers, that I am aware of, even though the law of diminishing returns should apply. It is important not to approach the minimum wage issue with theory alone, but with evidence about what happens in practice. Paul has pointed out some potential disadvantages to raising the minimum wage, but not the major advantages. Workers are also consumers. The more money they have in their pocket, the more they will spend, particularly in the local economy, which benefits local businesses. The evidence about the effect of the minimum wage is mixed. The following is a study of the relationship between the minimum wages and employment in all ten Canadian provinces between 1983 and 2012, which found that the minimum wage neither increased nor decreased employment: https://www.policyalternatives.ca/publications/reports/dispelling-minimum-wage-mythology; this is a University of California, Berkeley study about the potential effects of a $15 minimum wage in New York State, which takes into account both the positive and negative factors observed in other research of actual minimum wage increases: http://irle.berkeley.edu/files/2016/The-Effects-of-a-15-Minimum-Wage-in-New-York-State.pdf  

Thanks for your input, Marika.  I'm glad you pointed out that studies on the topic of the Minimum wage have been mixed.  That's why this is such an interesting topic of discussion, in my opinion. 

A recent study by the University of Washington in the United States found negative effects from the city of Seattle raising their Minimum wage too high.  That's what prompted me to ask this question.  

I think you might find this study interesting:

Link:  https://www.washingtonpost.com/news/wonk/wp/2017/06/26/new-study-casts-doubt-on-whether-a-15-minimum-wage-really-helps-workers/

Paul and others,

I will start with Paul's last point on the UW study which got a lot of press because its results were so out of line with all other studies. That same week, the Berkeley crew also released their study on the Seattle minimum wage with fairly conventional results (especially for those supporting the minimum wage), which got virtually no press. There are multiple reasons to suspect that the UW study is flawed methodologically, so I would not put much credence in it. Critiques of it by EPI (epi.org/130743) and Reich (CWED) in a letter to Robert Feldstein of the Seattle Office of Policy & Innovation detail many of the problems in the UW study. Anyway, while the popular press is going bananas over this study, I think that members of the profession will look at it differently. Although it is in an NBER publication, it is not peer reviewed. It will however be cited frequently, I'm sure.

To Paul's initial question regarding whether there is an upper limit on the minimum wage, I interpret that to mean whether there is some level of the minimum wage where the costs of raising it further exceed the benefits. (Note that that questions begs the question of which costs and benefits as well as how we would weigh them since some people would gain and others lose). On a conceptual level, it seems intuitive that at some level in a given time frame and market, that should be the case. In a way this is similar to the Laffer curve, that at some tax rate, increases in the rate will cause tax revenue to fall.  Having said that, the real question is at what level this occurs?

The other discussants here have enumerated various costs and benefits of raising the minimum wage. I would add to the costs that employers are less likely to offer general training (since employees can't "pay" for it with a lower wage during the training period) and the possibility of lowering or not offering other benefits to the job, i.e. replacing one form of compensation with another. However probably the most important thing accomplished by a higher minimum wage is to compress the wage scale.

Returning to why there must exist an upper limit. Most studies find that employment decreases at higher minimum wages and even those that don't (Marika's citation of the NYS study) say that if the wage was higher than 15 it is likely that employment would decrease. Employers will face the higher costs of the higher wage before the macro effects of higher incomes kick in. However, even in the best case scenario, employment will peak at some point and all that will result is higher inflation, i.e. no more gains but costs still increase. Of course it is likely that employment particularly of the unskilled will fall dramatically before then.

So I would conclude that there is an upper limit, however it will depend on the individual labor market in question and on the prevailing wages in that area, prices etc. as such, it will differ from place to place and over time and be very difficult to quantify.

Lastly, to Marika's point on CEO pay, I would similarly argue that an upper limit also exists and faces many of the same problems. The difference is that a minimum wage is putting a floor on wages overall while CEO compensation gets approved by the board (and ultimately shareholders who are paying it out of their own pockets) . However no one is setting a floor on CEO compensation (not that that is a problem).

Thanks for your input, Lee.  You make some great points!  

I agree with your conclusion.  There has to be an upper limit to the Minimum wage (where the negative effect outweighs the positive effect).  However, it probably varies with each geographical area and it would be difficult to determine exactly where that wage lies.  But I don't think its impossible to determine the upper limit in a specific geographical area if you have access to the right kind of data.  

I agree that conceptually with a wealth of data and some agreed upon criteria regarding comparing costs and benefits of gainers and losers (compensation principle perhaps), it could be done for a particular time and market. As a practical measure, I am not that optimistic. Again I turn to the Laffer curve as a comparison. There, the question is to find the tax rate that maximizes tax revenue, so that if we are at a higher tax rate we can lower the rate and increase revenue. That is merely a technical question, which I don't believe we have solved. Your question is harder because it also involves valuations. So I guess that I agree with you theoretically, but don't think we can get there empirically.

Good answer, Lee.  I think you're probably correct.  Since we haven't solved the Laffer Curve puzzle yet (i.e. the tax rate that maximizes tax revenue) then the maximum minimum wage in a specific geographic area is probably a larger challenge.  Good discussion!  I really appreciate your input since this is your area of expertise!  

The issue of the effects of MW on labor market is highly controversial, both in theory (competitive LM model versus monopsony) and in practice (Card & Krueger vs. Neumark & Washer). I think that the empirical consensus is that there are relatively small negative effects of MW on some groups of  low-skilled workers(e.g youth). Most existing studies, however, studied relatively small MW hikes, 20-30%. But what if we double MW in some country, region, or city? And this issue is close to the issue raised by Paul. Unfortunately (or fortunately) such examples of doubling MW are rare. What we found for Russia (https://link.springer.com/article/10.1186%2Fs40175-016-0051-0) is that the effects in fact are not dramatic. We found a very modest decline in youth employment and increase in informal employment as a reaction to MW hike. Of course, it is difficult to extend these conclusions to USA , (e.g., due to different strength of enforcement of MW legislature, tax morale, inflation rate, informal sector, etc) but still this suggest that the 'upper limit' is much higher than may be expected according to classic competitive LM model.

Minimum wage is a minimum floor level wage below which no employer can hire Labour . So with the increased inflationary pressure on prices, revision and increase in minimum floor level wages from time to time is necessary. But it has no upper limit. Upper limits are indicated by fair wages and decent wages. 

Minimum wage itself works as a supplement to social protection. I think MW doesnot have negative effects on employment, but will support the reduction to inequality  and increase in social justice. displacement of Labour is a natural process and it becomes higher in higher level of market wages.  Automation and use of more capital intensity is spontaneous whether MW is increased or not.

First, I want to thank Aleksey for making me aware of his article, which I plan to read. While the Russia case, where the minimum wage more than doubled, is an example of a major change  which virtually all papers exclude in their disclaimers, It is instructive that he finds only a modest decline in youth employment. However as he states "Of course, it is difficult to extend these conclusions to USA , (e.g., due to different strength of enforcement of MW legislature, tax morale, inflation rate, informal sector, etc)".. So the result is interesting and perhaps encouraging, but probably not useful as a guide to US or a state's policy. I do however look forward to reading it.

In regard to Umesh's comment, you are missing what I interpret to be Paul's point. He is not asking whether in all the potential futures, (and infinite time) whether the MW could continue to grow seemingly without end. Rather he is looking at a static model where for a given time and place, given the values of all the exogenous variables of the model e.g. price level, other wage rates, etc., dose there exist some maximum level of the MW where the costs of raising it further exceed the benefits (we would have to define both costs and benefits). We have addressed that question previously. I would add that while there are studies that show an increase in employment after an increase in MW, these positive effects are generally insignificant and are most likely not causal. Rather as Aleksey states " I think that the empirical consensus is that there are relatively small negative effects of MW on some groups of  low-skilled workers(e.g youth)." I would agree that to the extent that a consensus exists, that there are relatively small disemployment effects. Of Course there is a smaller consensus around Neumark's views that it is somewhat larger. Even most supporters of raising the minimum wage are reluctant to claim long run positive effects on employment (and certainly not for all classes of labor). If increases in the MW always led to higher employment (across all labor segments), then if $12 is good, $15 is better, and $10,000 is better still. That would make no sense and certainly even if true for some low levels of MW would not be true for higher ones. At some point, ceteris paribus, increases in the MW will not lead to increased employment, but rather lower employment, particularly for some groups.

Umesh is probably correct that one impact of raising the minimum wage is to compress the wage scale and consequently reduce income inequality. However if significant disemployemnt effects result, this would not be the case.

It may be of interest to you to be aware of a study that I initiated several decades ago with a model and results about the effect of the real minimum wage that I am not sure whether the literature that you look at considered.

The effect of an increase of the real minimum wage was studied within a model with TWO labor inputs: the number of employees and hours of work per employee. Our model assumed (a la Oi) that there is a cost to changing the number of employees but not for hours per worker. (More generally, for the hours the cost is lower than for number of employees). Within such a model an increase in the (minimum) has, among others, also an effect of substitution against hours per employee and in favor of the number of employees. The empirical results were that number of employees did not decline while hours per employee declined.

Within such a model the low skilled workers COULD be better off, and within our results they WERE better off 

The study did not have anything to do with an upper limit to a beneficial minimum wage.

The study was presented in a conference, but I do not think that it was published.

Thanks for that information, Shmuel.  That is very interesting.  I appreciate your input very much. 

Shmuel, thank you for your message. We also thought about this "substitution effect" that a MW increase could have. Could you outline the mechanism, why hours per employee should decline? Concerning Russia, the MW legislature covers only those workers who are employed full-time (depending on what is considered to be full-time at a particular job. For instance, doctors and teachers employed in public sector work less than standard 40 hours per week, but they are full-time employees). It implies that employers may adjust to a MW increase by moving some workers from full-time to part-time positions, and these workers will be not covered by MW. As a result, hours per worker decline but employment does not.

Sorry for the delay, Aleksey.

In the model with one type of labor, number of low-skilled employees E, and capital K, the implicit assumption is that hours of work per employee H, cannot be changed, e.g. by law. A higher minimum wage cause labor to become more expensive, causing marginal costs to rise, output to decline and (normally) employment to decline (output effect both in the "short run" and in the "long run"). The substitution effect is against employment and in favor capital in a "long run".

With 2 labor inputs, E and H, when H can be changed at no costs to the employer while E can be changed with a cost, e.g. hiring and training cost a la Walter Oi (JPE 1961, I think) or severance payment when laying off. The cost function is now wHE + eE + kK. The price of (changing only) H is wE, while the price of employment E is wH + e. Raising the (minimum wage) w by 10% will raise the price of H by 10%, but the price of E will rise by LESS than 10%. E become relatively cheaper compared to H, causing a substitution effect between them in favor of E.

H is predicted to decline due to an decline in output as well as due to a negative substitutions with E and with K. For E, the outcome is less predictable. If the positive substitution for E with H is large enough, to more than compensate for the negative substitution of E with K and the negative output effect for E, the number of employees may not decline; E may even increase.

The effect on wH, the income of a worker, is not predictable as w rises and H declines. The same is true for wHE, the total cost of labor, but since E may increase, total labor income may increase even if the income of current employees will decline. The likelihood that wHE will rise, increases if more employees are hired.

An individual current employee that does not lose the job, IS better when his/her income rises; but even if her/his income falls, he may still be better off. The decline in H means more leisure for the employee, and as long as the marginal utility of leisure is positive, it may compensate the worker enough for the loss of some income.

If E falls, the laid off employees are worse off as in the model with one type of labor, E.

The expanded model suggest many possibilities of doing good by raising the minimum wage compared to the more simple model.

I hope that the model and its predictions are clear. I hope that Aleksey, or anyone else, will want to pursue the expanded model. It is important, I think, to test the predictions empirically in different countries, while holding capital fixed ("short run") or when it is changing ("long run").

If anyone will be interested enough, I may look for my unpublished material, to see how much I pursued the topic. I am almost sure that I have done something empirical as well. 

Best wishes to all.

Paul, contrary to my first comment, I think that the model with the two labor inputs (see my response to Aleksey today) MAY provides some theoretical response to your question. If, as a result of an increase in the minimum wage there is a reduction in H (hours per worker), in E (Number of employees) and in wH (income of an employed worker who was NOT laid off), then the likelihood that will be workers who become better off is small. 

The only issue is whether there is a way to show that workers who are not laid off but suffer an income loss are not compensated enough by their increased leisure time. Can someone deal with the issue?

If, and this is a BIG IF, the point can be proven, then there will be no rationale to raising the minimum wage so high. Even then, this will be a theoretical maximum, and empirical studies will be needed to get a relevant number in countries, industries, period, etc.

Since the scientific revolution led to the industrial revolution humanity has experienced an explosion of production and wealth the likes of which have never before been seen on this earth.

Virtually EVERYONE lives better than most kings lived during much of history.

The kings of the middle ages did not speed down the street in horseless chariots, heated and air conditioned ... with symphonies at their fingertips, they could not cure even the simplest diseases or get fleas out of their own beards.

The great concern of western governments has passed from manipulating laws and behaviors to enable their people to PRODUCE more wealth to the redistribution of what appears to be endless abundance.

Government is increasingly preoccupied with proving everyone roads, schools, healthcare, food, homes, and minimum wages.

When the day comes that only one of the world's robotic factories produces all the cars needed by everyone ... and only one other factory produces all the shoes ... and only one can produce all the phones, etc ...

What social-systems will need to be in place to ensure all this wealth is distributed to everyone around the world?

Our problem will no longer be wealth creation, but wealth distribution.

And we may be in the transitional period between these two eras of history right now.

This is discussed at length in a book titled Our Human Herds

www.ourhumanherds.com

The world described by Stephen Martin Fritz is still not different from ours when it comes to the problem of wealth creation. The law of scarcity will still apply. There will still be limited resources and alternative costs will exist for any project of production. Thus, even in that world there will be a problem of wealth creation.

I am sure that at some point a higher minimum wage will have a negative effect on the workers and the employers, i.e. no one will be a winner. I do not know what is that level.THIS IS AN EMPIRICAL, NOT THEORETICAL ISSUE.

This is not a problem unique to minimum wage. The Laffer curve Correctly suggests that raising the tax rate from 70% to 100% tax rate, the tax revenue will fall to zero if people are not required to work. Will that happen at 90% to 91%, or at 19% to 20% we do not know. We know that lowering rates from around 40% to 20% tax revenue FALL since this is what happened to Reagan (in the 1980s and to Trump in 2017/9.

Minimum wage is a minimum floor level wage below which no employer can hire Labour . So with the increased inflationary pressure on prices, revision and increase in minimum floor level wages from time to time is necessary. But it has no upper limit.