A plot tries to answer the questions like What kinds of skills H-1B workers need to survive economic crisis or How cross-functional skills contribute to foreign employees' wage.
The figures on the left, are the employment rate from 1999 to 2018, extending over two well-known economic crisis i.e. the 2000 Dot-com Bubble and the Financial Crisis of 2007-2008. The figures on the right are the sub-industries under the individual industries. Bars show the annual wage (in U.S. dollar) of the H-1B workers fall into the sub-industries. Stacked colors are shown in proportional to the cross-functional skills that required for the occupation.
The 9 selected industries are Management and technical consulting services, Computer systems design and related services, Scientific research and development services, Legal services, Accounting and bookkeeping services, Advertising and related services, Specialized design services, Architectural and engineering services and Other professional and technical services, all belongs to Professional, Scientific, and Technical Services (NAICS code 54 [1]). The employment rate data are collected in month precision, on non-supervisory employees or production workers nation-wide.
Cross-functional skills are defined as developed capacities that facilitate performance of activities that occur across jobs:
Developed capacities used to solve novel, ill-defined problems in complex, real-world settings
Developed capacities used to design, set-up, operate, and correct malfunctions involving application of machines or technological systems
Developed capacities used to understand, monitor, and improve socio-technical systems
Developed capacities used to allocate resources efficiently
In this research, we analysed data from Current Employment Statistics (CES) by Bureau of Labor Statistics, H-1B Case Disclosure Data by Office of Foreign Labor Certification (OFLC) , Occupation skill set data by O*NET and North American Industry Classification System (NAICS), revealed the some relationships in un-/employment rate versus some phenomenal historical events (financial crises), wage level and cross-functional skills [2] for occupations in selected industries across finance, legal, computer and etc.
The 9 selected industries (from Professional, Scientific, and Technical Services) shows diverse patterns in employment growth rate over the decades. It's worthy of mention that the year 2000 and the year 2008 cause significant negative employment rate on most industries including advertising, accounting, architecture and engineering, specialized design and etc, which may have strong correlation to the 2000 Dot-com Bubble and the Financial crisis of 2007–2008.
Computer system design and its sub-industries have better resistance against the 2008 mortgage crisis, however it was struck hard in the 2000 crisis. The legal shows no negative growth in the 2000 crisis but like most industries, had a huge drop in employment rate in 2008.
Interestingly, the scientific research and development shows no correlation to the two financial crises. The sub-industries included
US Bureau of Labor Statistics | Estimates of employment, hours, and earnings information on a national basis. [KAGGLE]
The Office of Foreign Labor Certification releases program disclosure data on a quarterly basis on the OFLC Case Disclosure Data page which provides H-1B disclosure data. [OFLC] [KAGGLE]
National Center for O*NET Development | Provide a mapping of O*NET-SOC codes (occupations) to Skill ratings. [O*NET]
U.S. CENSUS BUREAU | NAICS is the standard used by Federal statistical agencies in classifying business establishments for the purpose of collecting, analyzing, and publishing statistical data related to the U.S. business economy. [census]
The national employment rate data is aggregated from 620 detailed industries, over 216 months (18 years). First, the data is grouped by the industries at a large industry sector hierarchy (the industry of NAICS 54 - Professional, Scientific, and Technical Services), and then for each series of data we aggregated the employment data index (data type 06 - Production or nonsupervisory employees) over the 2 decades, month by month. Next the employment rate is calculated by the difference of the adjacent months then divided by the former month. To show a smooth plot, we leverage quadratic interpolation on time-axis, producing a smooth, organic looking style. For more information on the data pre-processing of this part, please also refer to my INFSCI 2415 Assignment I.
The skill-wage bar plot is a bar plotting on each individual sub-industry. For each industry, H-1B employment data is grouped and aggregated by the median of the annually wage. For a simple wage model, we dropped all data row that was not statistics annually (e.g. hourly, monthly, semi-monthly wages). The industry-wised aggregations were possibly via the NAICS code.
The skill stacked plot is produced with the cross-functional skill score data in 5 metrics: Social Skills, Complex Problem Solving Skills, Technical Skills, Systems Skills and Resource Management Skills. O*Net cross-functional scores are measured individually at the scale of \(5.00\). To visualize the sense of skill stacked on wage, we sum the 5 scores and scale all the sum scores to the annual wage level while keeping the skill score proportional to the original scores, as:
Where \(C_i\) is the proportional score of each cross-functional skill;
\(W\) is the median wage of the sub-/industry;
\(s_1\): Social Skills;
\(s_2\): Complex Problem Solving Skills;
\(s_3\): Technical Skills;
\(s_4\): Systems Skills;
\(s_5\): Resource Management Skills.
Also checkout the
Jupyter Notebook file for this part.
Economic crisis is almost inevitable.
The root cause of the great depression was the development of technology, which led to the increase in productivity, and then the domestic market could not meet the demand for industrial production, leading to mass unemployment and depression. Mass unemployment would result, however, because people living in the cities would not have the means to survive without an income, and they would be in danger of dying (unlike the farmers, who at least could feed themselves).
While for every country, productivity growth is a dream, which means a rise in international competitiveness, so that neither country nor company can produce too much and hold back technology. Moreover, the development of science and technology is not subject to the will of the government.
So not only is the bargaining power of workers low, but society must provide enough jobs or there will be social unrest (in the shadow of hunger and death, it is understandable for the unemployed to do anything). In other words, the industrialized countries cannot afford serious unemployment.For most people, the most important way to participate in the distribution of wealth is through employment. For h1-b workers, the main way to stay in the U.S. is to work. Unemployment is a nightmare for them.
In the 70 years since the end of the second world war, the United States has had 11 economic crises, one every six years on average. For h1-b workers, it's been one disaster after another. For them, it will take more than most Americans to avoid elimination in the great depression.As you can see from the top right, we have nine broad categories‘ composition of cross-functional skills. It can be learned that the requirements for employees in various industries are no longer limited to their professional abilities, but involve more comprehensive industries. This also means that in order to ensure their professional level to remain competitive, they need to have more comprehensive capabilities. The best way to predict the future is to create it. That is, by thinking about, discussing the possibilities of the future and taking action to avoid, they can create the best model of the future to move forward in.
Authors: Guangxue Wen | Guangqin Ran | Ping-Kang Huang
INFSCI 2415 Information Visualization | Prof. Lingfei Wu