One chart shows the employment impact of Seattle minimum wage increase

[Axulus]

Still think increases in the minimum wage don't affect employment?

The 2015 decrease happened during time of strength in Seattle's economy:

As of February, total nonfarm employment for the Seattle-Bellevue-Everett metropolitan division reached nearly 1.6 million, a record. The velocity is as impressive as the late 1990s, although it hasn’t run as long (yet).

http://www.seattletimes.com/business/economy/perfect-mix-paved-way-to-boom/

 

[AthenaAwakened]

Fox News is hyping a report from the conservative American Enterprise Institute (AEI) blaming a marginal decline in restaurant employment in the Seattle-Tacoma-Bellevue metropolitan area on Seattle's recently-increased minimum wage. The think tank and right-wing media outlet both overstated the significance of a roughly 1 percent change in restaurant employment and focused on apparent job losses in one month while ignoring job gains the following month.

http://mediamatters.org/research/2015/08/11/fox-hypes-cherry-picked-data-to-attack-seattle/204889

Decline In Restaurant Employment Is Less Than 1 Percent. According to data from the Bureau of Labor Statistics (BLS) and compiled by the Federal Reserve Bank of St. Louis, total employment in the combined MSA in "food services and drinking places" declined from 135,300 to 134,000 from January through June 2015. This 1,300 reduction, which is still subject to revision, represents a less than 1 percent reduction in the area's restaurant workforce:

[Federal Reserve Bank of St. Louis, accessed 8/11/15]

Total Employment Continues To Climb Despite Wage Increase. Beyond the restaurant industry, total employment in the region increased by nearly 22,000 from January through June 2015, according to data from the BLS. City, state, and federal minimum wage laws in this case would apply to all businesses, not just those in the food service industry:

[Federal Reserve Bank of St. Louis, accessed 8/11/15]

Seattle Times: Claim That Restaurant Struggles Are "Linked" To Seattle Wage Increase Are "False." According to a March 17 report by The Seattle Times, business and restaurant owners in the city expressed little anxiety about the upcoming April 1 minimum wage increase. The paper interviewed numerous business owners and found claims that the increase would hurt restaurants to be "false":

 

[Jimmy Higgins]

OMG! Axulus posted somebody else's hatchet job which would have only required a couple of minutes to debunk. Shocked!

See! See! 1000 jobs lost. You increase the wage to $15 and 1% of people lost their jobs! We told you, we told you, armageddon and 1% job loss!

 

[bilby]

Still think increases in the minimum wage don't affect employment?

The 2015 decrease happened during time of strength in Seattle's economy:

As of February, total nonfarm employment for the Seattle-Bellevue-Everett metropolitan division reached nearly 1.6 million, a record. The velocity is as impressive as the late 1990s, although it hasn’t run as long (yet).

http://www.seattletimes.com/business/economy/perfect-mix-paved-way-to-boom/

The April gain of nearly 4,000 is clearly an outlier, and it would be reasonable to expect the following months to show a decline as a result.

Just by eye, it seems that the long-term mean is between 1,000 and 2,000 per month outside recessions; the May and June figures are entirely consistent with the correction of the outlier in May. There is nothing about the June figure that needs to be explained by factors other than those presented directly in the graph itself.

 

[Axulus]

Still think increases in the minimum wage don't affect employment?

The 2015 decrease happened during time of strength in Seattle's economy:



http://www.seattletimes.com/business/economy/perfect-mix-paved-way-to-boom/

The April gain of nearly 4,000 is clearly an outlier, and it would be reasonable to expect the following months to show a decline as a result.

Just by eye, it seems that the long-term mean is between 1,000 and 2,000 per month outside recessions; the May and June figures are entirely consistent with the correction of the outlier in May. There is nothing about the June figure that needs to be explained by factors other than those presented directly in the graph itself.

That chart isn't per month. It's the change in restaurant jobs for the first 6 months combined of each calendar year going back to 1990.

 

[Axulus]

OMG! Axulus posted somebody else's hatchet job which would have only required a couple of minutes to debunk. Shocked!

See! See! 1000 jobs lost. You increase the wage to $15 and 1% of people lost their jobs! We told you, we told you, armageddon and 1% job loss!

Graph and data = hatchet job - thank you for your contribution.

By the way, the minimum wage isn't yet $15. It is being phased in over a few years. This is just phase one.

But I guess it's ok if a few thousand "undesirables" are now unemployed and on the dole (normal trend was ~2,000 job gains vs. 1,200 losses, for net anomaly of 3,200). Fuck them anyway. They must not have contributed anything worthwhile to society if they can't manage what it takes to make the new "living wage".

 

[bilby]

The April gain of nearly 4,000 is clearly an outlier, and it would be reasonable to expect the following months to show a decline as a result.

Just by eye, it seems that the long-term mean is between 1,000 and 2,000 per month outside recessions; the May and June figures are entirely consistent with the correction of the outlier in May. There is nothing about the June figure that needs to be explained by factors other than those presented directly in the graph itself.

That chart isn't per month. It's the change in restaurant jobs for the first 6 months combined of each calendar year going back to 1990.

Well that's a pretty bizarre dataset. The absence of the second half of each year pretty much renders the dataset valueless for any purpose - and makes me suspect that someone is cherry-picking their data. When half the data is not shown, my first thought is 'what is in that half of the data that they are concealing from me?'

For a continuous timescale, regardless of the granularity of the data, the analysis I gave remains the same - a much larger than usual increase should be expected to be followed by a compensating decline. The omission of half of the data for each year does make a difference though - it means that no valid expectation can be applied to any of this information, and that the graph is completely useless. The July-December data are essential in order to make any sense at all of the information presented.

You might as well have shown a graph of the half-time scores in Seattle Seahawks matches over the last five years - that would tell us exactly as much about the effect of MW on employment as the graph you actually presented.

 

[Axulus]

That chart isn't per month. It's the change in restaurant jobs for the first 6 months combined of each calendar year going back to 1990.

Well that's a pretty bizarre dataset. The absence of the second half of each year pretty much renders the dataset valueless for any purpose - and makes me suspect that someone is cherry-picking their data. When half the data is not shown, my first thought is 'what is in that half of the data that they are concealing from me?'

For a continuous timescale, regardless of the granularity of the data, the analysis I gave remains the same - a much larger than usual increase should be expected to be followed by a compensating decline. The omission of half of the data for each year does make a difference though - it means that no valid expectation can be applied to any of this information, and that the graph is completely useless. The July-December data are essential in order to make any sense at all of the information presented.

You might as well have shown a graph of the half-time scores in Seattle Seahawks matches over the last five years - that would tell us exactly as much about the effect of MW on employment as the graph you actually presented.

Hmm, a law that went into effect towards the beginning of the year compares the effect of the law up to the current point that data is available (6 month time period), and then compares that to prior years for the same period with 25 years worth of data and shows a clear anomaly. Very suspect...

Does a statistician here want to run the p-value on this data set in regards to the null hypothesis that the minimum wage law had zero effect on employment? I'll place a large bet that this is a statistically significant result (p value less than .05)

 

[bilby]

Well that's a pretty bizarre dataset. The absence of the second half of each year pretty much renders the dataset valueless for any purpose - and makes me suspect that someone is cherry-picking their data. When half the data is not shown, my first thought is 'what is in that half of the data that they are concealing from me?'

For a continuous timescale, regardless of the granularity of the data, the analysis I gave remains the same - a much larger than usual increase should be expected to be followed by a compensating decline. The omission of half of the data for each year does make a difference though - it means that no valid expectation can be applied to any of this information, and that the graph is completely useless. The July-December data are essential in order to make any sense at all of the information presented.

You might as well have shown a graph of the half-time scores in Seattle Seahawks matches over the last five years - that would tell us exactly as much about the effect of MW on employment as the graph you actually presented.

Hmm, a law that went into effect towards the beginning of the year compares the effect of the law up to the current point that data is available (6 month time period), and then compares that to prior years for the same period with 25 years worth of data and shows a clear anomaly. Very suspect...

Does a statistician here want to run the p-value on this data set in regards to the null hypothesis that the minimum wage law had zero effect on employment? I'll place a large bet that this is a statistically significant result (p value less than .05)

The clear anomaly on the graph is in the third period from the right - a value of 4,000 is a massive outlier. Do you contend that the minimum wage introduced two years after this figure is responsible; or do you accept that either there are other factors far bigger than any MW effect that are in play; or that the entire dataset is meaningless and no conclusions drawn from it are valid?

No significant information can be gleaned from half a dataset. If you had a graph showing the July-December figures interspersed with the January-June numbers - ie showing EVERY six month aggregate figure, rather than every second six month figure - then you might be able to draw some kind of inference from it. But you haven't shown such a graph; and the graph you have shown is valueless with half the data missing.

The only reason I can think of for showing HALF the data, and not the other half, is to conceal information that does not support the agenda of the person who made the graph. Of course, it is possible that they are just too stupid to realise that they have rendered their graph useless; It is perfectly possible that the full dataset does support the conclusion they are drawing. But the point remains that they cannot get to that conclusion from this dataset. You might as well have shown a graph of the half-time scores in Seattle Seahawks matches over the last five years - that would tell us exactly as much about the effect of MW on employment as the graph you actually presented.

Note that my main objection comes BEFORE any analysis - the dataset is incomplete and any analysis of it is therefore valueless. You can discuss statistical significance until the cows come home - a cherry picked dataset that does NOT show significance at the (arbitrary) p<0.05 level would be a surprise, and would indicate truly staggering incompetence on the part of the propagandist who made the chart.

It is the demonstrable fact that not all of the data is present that matters. Analysis of the broken dataset actually presented is MEANINGLESS.

 

[bleubird]

Why would anyone begrudge a person a living wage to feed you. If you are too busy or too lazy to feed yourself,than you should be willing to pay a fair price for that service.
Food service is hard.Both for the worker and the owner.
I think all high school students should work at lest one month at a food service job.

 

[laughing dog]

Well that's a pretty bizarre dataset. The absence of the second half of each year pretty much renders the dataset valueless for any purpose - and makes me suspect that someone is cherry-picking their data. When half the data is not shown, my first thought is 'what is in that half of the data that they are concealing from me?'

For a continuous timescale, regardless of the granularity of the data, the analysis I gave remains the same - a much larger than usual increase should be expected to be followed by a compensating decline. The omission of half of the data for each year does make a difference though - it means that no valid expectation can be applied to any of this information, and that the graph is completely useless. The July-December data are essential in order to make any sense at all of the information presented.

You might as well have shown a graph of the half-time scores in Seattle Seahawks matches over the last five years - that would tell us exactly as much about the effect of MW on employment as the graph you actually presented.

Hmm, a law that went into effect towards the beginning of the year compares the effect of the law up to the current point that data is available (6 month time period), and then compares that to prior years for the same period with 25 years worth of data and shows a clear anomaly. Very suspect...

Does a statistician here want to run the p-value on this data set in regards to the null hypothesis that the minimum wage law had zero effect on employment? I'll place a large bet that this is a statistically significant result (p value less than .05)

Come on, you have to admit that there are many influences on the volatile restaurant trade: one data point with controlling for other factors is not terribly convincing of anything. And I think that the minimum wage hike probably did contribute to some loss in jobs.

 

[untermensche]

Where's the other chart?

The one showing the happiness levels in people working in the restaurant trade.

Oh yeah. Life is not about human happiness.

It is about getting rich and exploiting others.

 

[Jimmy Higgins]

OMG! Axulus posted somebody else's hatchet job which would have only required a couple of minutes to debunk. Shocked!

See! See! 1000 jobs lost. You increase the wage to $15 and 1% of people lost their jobs! We told you, we told you, armageddon and 1% job loss!

Graph and data = hatchet job - thank you for your contribution.

It is a data point, nothing more.

By the way, the minimum wage isn't yet $15. It is being phased in over a few years. This is just phase one.

But I guess it's ok if a few thousand "undesirables" are now unemployed and on the dole (normal trend was ~2,000 job gains vs. 1,200 losses, for net anomaly of 3,200). Fuck them anyway.

Exactly. Armageddon, the sky is falling, we told you huge job losses... err... maybe a 1,000.

They must not have contributed anything worthwhile to society if they can't manage what it takes to make the new "living wage".

Nothing like bullshit spin. So how many workers wages were increased verses the number of labor positions that disappeared?

 

[Nice Squirrel]

But does this really tell the whole story. The small decrease could* be accounted for an exit of cooks to better paying jobs. (We have a shortage of cooks in our city and the prevailing wage is $15/hr.)

 

[Loren Pechtel]

Still think increases in the minimum wage don't affect employment?

The 2015 decrease happened during time of strength in Seattle's economy:



http://www.seattletimes.com/business/economy/perfect-mix-paved-way-to-boom/

The April gain of nearly 4,000 is clearly an outlier, and it would be reasonable to expect the following months to show a decline as a result.

Just by eye, it seems that the long-term mean is between 1,000 and 2,000 per month outside recessions; the May and June figures are entirely consistent with the correction of the outlier in May. There is nothing about the June figure that needs to be explained by factors other than those presented directly in the graph itself.

Check that graph again.

 

[dismal]

But does this really tell the whole story. The small decrease could be accounted for an exit of cooks to better paying jobs. (We have a shortage of cooks in our city and the prevailing wage is $15/hr.)

And yet (some) people must fight tooth and nail against the argument based in both common sense and basic economics that "when something costs more people use less of it" instead of arguing this.

 

[Nice Squirrel]

basic economics that "when something costs more people use less of it"

Yes, this is a really dumb argument, but typical for someone who never got beyond "basic economics".

 

[dismal]

basic economics that "when something costs more people use less of it"

Yes, this is a really dumb argument, but typical for someone who never got beyond "basic economics".

I took graduate school economics and they never reversed that whole "when something costs more people use less of it" rule.

 

[SimpleDon]

Still think increases in the minimum wage don't affect employment?

The 2015 decrease happened during time of strength in Seattle's economy:

As of February, total nonfarm employment for the Seattle-Bellevue-Everett metropolitan division reached nearly 1.6 million, a record. The velocity is as impressive as the late 1990s, although it hasn’t run as long (yet).

http://www.seattletimes.com/business/economy/perfect-mix-paved-way-to-boom/

Is this now your standard for judging economic policy changes? The impact on employment?

Considering that it appears to be how can you justify deregulation and the failure to enforce existing regulations of the financial markets that caused the Great Financial Crisis and Recession of 2008 that threw 25 millions people out of work?

How can you justify austerity as an acceptable means to recover from the GFC&R of 2008 when the employment growth since the GFC&R has anemic compared to other recoveries primarily because of the unprecedented drop in government spending that is the result of the application of austerity?

Or for that matter how can you justify the continued application and doubling down of the neoliberal, supply side economic policies that have so clearly failed to deliver the growth in the economy and in employment that was promised for the policies?

I don't care if these are policies that you support. The question here is why you aren't posting anything in opposition to these job killing policies?

These are policies that have resulted in employment losses in the millions. How can you ignore them?

As for the situation in Seattle I have been consistent in saying that while I don't believe that raising the minimum wage will result in the widespread dis-employment* that neoclassical economics predicts for it, which depends on the rather dubious theory of marginal productivity, or for that matter the increase in employment that many others predict, which depends on the rather optimistic application of economic multipliers. That the overriding economic need for raising the minimum wage is to raise wages across the board, to shift the distribution of income from the wealthy to the the poor and the middle class. To reduce profits and to increase wages. To increase demand and to decrease supply.

That if this does result in dis-employment that it must be treated as an separate problem, not as a reason to continue to keep increasing income inequality, to keep intentionally redistributing income to the wealthy as conservatives and all of the Republican candidates are proposing.



* dis-employment is a combination of all of the failures of employment, unemployment, the conversion of full time work into part time work, the failure of employment to keep up with the growth in the population, etc., in general, the failure to provide employment to those who need it.

 

[ksen]

Hold on, why were there job increases at all after 2010? I thought Obamacare was going to kill all those jobs?