September 19, 2019
Brian W. Dew, @bd_econ
Various blog posts written by Brian Dew for the bd econ blog or for other organizations or publications.
Blog posts for bd econ blog or done for other organizations/publications
September 19, 2019
Brian W. Dew, @bd_econ
Various blog posts written by Brian Dew for the bd econ blog or for other organizations or publications.
I'm curious what share of workers in western MD and the eastern shore make less than 15/hr compared to Montgomery county or to Baltimore. Will look at a few MSAs and see how EPOP, unemployment, and share of workers paid less than 15/hr compares.
Some JOLTS results suggest labor market still not tight, others suggest it is getting very tight. Look into these.
Ideally with a single notebook that makes it replicable and using ACS data directly.
There appears to be enough info for a pretty nice summary of economic data that could foretell a recession.
Specifically, 1) wage growth rates and employment rates are turning over. 2) the economy is top-heavy and has been driven by consumer spending of people with high incomes. 3) Asset prices seem elevated and write-downs make sense given smaller expected streams of profit. 4) Health care and education costs are eating the disposable income of the bottom 80% and will reduce the ability of the Fed to use lower interest rates to boost spending.
Last couple months of CPS have very low response rate (49,500 households). Show trend in refusal rate. Also (try) to map refusal rate from previous 2-3 years--map will be complicated but good example of working through CPS geography issues. So if CSA, map CSA, if not map non-CSA state-level. If that doesn't yield sufficient sample, just map state.
Write up some findings from this data.
Write a post looking at how rural area ownership income is distributed, based on the March CPS.
Use four years of data to get a sample that allows looking at UNEMPTYPE and UNEMPDUR.
Look at GDP growth rates and underlying components since 1990 or so for the US, euro area, and Japan (and if possible the Nordics).
Look at time series for all-age epop. Then look at what percent of individual households (or families?) have positive net workers.
Write up an overview of what the bd CPS is, why I created it, and how to use it.
Look at what share of men and women, age 25-54 are Employed, Unemployed, NILF. Perhaps find a way to compare 2015-16 with 2017-18.
Bin the CPS ASEC families by net workers and calculate how much income (in total USD) is needed to move every person in each bin to the poverty line.
Remove the plotly pieces from the construction shortages notebook and make sure it still runs. Add a link to the actual blog post.
Should be interesting to see what percent of 21-23, 24-26, and 27-29 year olds (men and women separately) live with a parent.
Look at category of work, and common major occupations and major industries, compared to the US.
How close was the US to full employment in November 2019 to February 2020?
Tons to look at: LFS
, NILFREASON
, DWTYPE
, UNEMPTYPE
, PTECON
, HRSACTT
, ABSTYPE
, ABSPAID
, etc.
One chart with two lines since 2000: 1) USD difference between RHRWAGE for HS and for COLL; 2) USD difference between RHRWAGE for HS and [SC, COLL, ADV].
Write up a guide with examples for creating a small summary of the employment situation, by state, from the last twelve months of CPS data.
Add a post covering how many hours people usually work (average and median) and what share of people work more than one job.
Now that I have the full 2018 set of monthly CPS microdata files, write up some data on six southern labor markets: Asheville, Chattanooga, Knoxville, Kingsport, Greenville, and Huntsville.
I want to know about: age, family structure, education, industries, occupations, hours worked, NILF - retired, NILF - disabled, Epop, Epop with care (by gender), Median Wage, 10th percentile wage, unemployment, union membership, professional certification.
Perhaps turn this into a series of several blog posts. Start today with one of the above.
Write up some of the amazing data on oil production in New Mexico (from EIA). The state has a permanent fund and other funds that invested earlier oil revenue and pay into various state programs (primarily education) using the proceeds from the investments. The state has something like 2 million people living in an area equivalent in size to Poland. The oil production is in the southeast part of the state. Yet I recall (would need to confirm) that many people in the state, and even in the oil-producing region, are very poor. So it would make sense to look into whether there are new inflows in the various state funds and to see what the state budget looks like.
Check whether inflation is overstated for low wage workers.
My suspicion is based on a few observations:
Real wage estimates tend to deflate prices using the all metro area CPI
The regional CPI shows higher rates of inflation in the northeast and west
Low-wage workers tend to be in the south and midwest (places where the local minimum wage is more likely to be the federal minimum wage)
Therefore, measures of real wages at the bottom of the wage distribution may be artificially low.
To check this, calculate the real wage at the first decile using the CPI-U and again using the regional CPI-U.
Two possible graphs:
Share of bottom 10% of wage earners living in the south or midwest.
First decile real wage showing two lines, one with each deflator.
Then write up the results.
Look at epops in metro areas since 2013 (or whenever codes changed). Any patterns over time in the distribution? Are the low epop areas catching up?
If I'm correct, it would only take two sets of charts to point out the flaw in this paper. First, the six charts except showing investment share of GDP, second the six charts except showing IMF forecasts for some employment measure, to show that the employment miscalculation, not the TCJA is why their US forecast was so wrong.
Last post in the series will look at what percent of each area's workers are union members and separately what percent are covered by a union contract. Then look at what percent have some professional certification or license.
Median real weekly earnings for people in the six areas, compared to the US as a whole.
What share of each area is in each of the WBHAOM groups? What share is born outside of the US? What are the most common countries of origin for people born outside the US.
Look at educational breakdown, including by gender. Look at school enrollment for younger groups.
Summarize findings from the 10 posts.
Compare across several measures. Try to dig into hypothetical by decomposing GDP growth into total economy productivity, hours, population, epop for both countries. Also, standardize for population growth and compare GDP growth decomposed into total private and government consumption (ideally with health care and education set aside), net exports, investment (total, government and private).
There's some evidence that the CPS redesign in 1994 created problems for the unemployment duration variable. Usually, this is discussed from the perspective of short-term unemployment, but check whether it also affects the long-term share of unemployed.
Basically, the issue is that the CPS interviewer stopped asking unemployment duration for people unemployed the previous month (and just code it as the previous month value plus 4 weeks). This doesn't apply to MIS in [1, 5] so that should be a much better sample for measuring unemployment duration.
If I don't get anything, just ignore it. If there is some difference, even small, write it up.
Why are people not participating in the labor market in each of the six areas, compared to the US as a whole.
Trends around summer jobs and breakdown by COW, IND, OCC.
Are teachers more likely to work second jobs (using CPS matching--basically want to know about summer different jobs) over time and compared to others.
Write up some comments on a nationwide severance policy as a way to protect workers from disruption. Look into other country experiences.
Check the list of notebooks in various locations on the blog and github pages site. Fix any broken links and add a few new ones.
A clever test of whether an industry faces a labor shortage is checking if the industry wage growth outpaces industry productivity growth. This could be summarized well in a range bar chart that shows the gap between wages and productivity by industry (5 year high, five year low, average, latest).
Future charts in the southern six series should use this technique: https://matplotlib.org/3.1.1/gallery/lines_bars_and_markers/horizontal_barchart_distribution.html#sphx-glr-gallery-lines-bars-and-markers-horizontal-barchart-distribution-py
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.