Grammar of graphics

Image by pealco via Flickr

The grammar of graphics is a fine attempt to teach analytic graphics. Developed for the freeware statistical software R, the ggplot2 module is indeed a fine tool to use graphs to analyze data.

 Hadley Wickham has announced that new versions of his popular grammar-of-graphics charting package ggplot2 and his general-purpose data reshaping tool plyr for R are now available. 

Some graphs created by ggplot2 are indeed work of art, especially the ones where multivariate data are presented in a single graph, such as the histogram matrix.

Microsoft Excel while made charting easy for those who were not empirically inclined, however, the 3-D graph options and the like made way for certain very ugly and misleading graphs. Nathan Yau's Seven Basic Rules for Making Charts and Graphs is indeed a good starting point for those who are itching to use graphs, but do not know where to begin.

Related articles by Zemanta

• 7 Basic Rules for Making Charts and Graphs (flowingdata.com)

Zemanta: Bloggers’ assistant

There are times you wish you had an assistant to provide you with the facts and data as you are busy writing your blog. Zemanta offers a tentative solution. While you write your blog, it tries to provide you with the background material.

Let’s take Zemanta for a test drive. I would like this blog about the rising debt levels amongst Canadians. A recent Bloomberg story highlighted the fact that Canadians are getting more and more into debt. That is the ratio of debt to  disposable income for households have been on the rise.  If these trends continue, Canadian households are likely to pullback spending in the months to come.

The household debt has been rising in the US as well. See the image below that suggests that the household debt in the US has peaked.

According to an entry on Flickr:

On December 10th, the Federal Reserve released its latest  latest findings on Consumer Debt Outstanding. In reflection of good personal choices, personal debt went down by 2.6% for the third quarter 2009. This is amazing because it brings the total personal national debt down to around $13.6 trillion. Also, for the first time ever, personal debt did not grow for the fifth quarter in a row. This decrease was due to a 13.6% drop in household home mortgage debt AND a 3.2% drop in consumer credit.

 

Image by eric731 via Flickr        

How did I get this image? Zemanta suggested it.

What about other related stories and information. According to Windsor Star, Canada’s household debt has reached $1.4 trillion. I didn’t search for this. Zemanta suggested it. By the way notice the 10% rule. If the US households are carrying $14 trillion in household debt, their Canadian counterparts would carry 10% of the US amount.

Another story in Toronto Star on May 11 refers to the fact that Canadian households are the most indebted people living in the advanced countries, according to OECD.

How do I know this? Zemanta suggested it.

Lastly, the Bank of Canada's governor offered the following comment on household debt in Canada in a Reuters' story:

"We don't expect a sharp uptick in household savings in Canada. Again, that is one of the risks to the projection that there could be a repair of household balance sheets, or an increase, I should say, in household savings in Canada. All of that said... household balance sheets in Canada are in quite strong shape and that is one of the big differences between Canada and the United States."

Who searched this item for me? Zemanta did.

Related articles by Zemanta

• Household debt skyrockets (windsorstar.com)
• Canadian households among the most indebted of OECD (thestar.com)
• HIGHLIGHTS-Bank of Canada's Carney speaks in Ottawa (reuters.com)

The Econometrics bestsellers

A look at the global bestsellers (sold through Amazon) returns the usual suspects for Econometrics. If you are interested in learning about the top 10 best selling books in econometrics in the US, Canada, England, Germany, and France, click HERE. The top-3 bestsellers in July 2010 in the US, Canada, and UK are presented in the image below.

There is an interesting similarity in the choices made by consumers located thousands of miles apart. Consider that Mostly Harmless Econometrics: An Empiricist's Companion by Joshua D. Angrist et al. is listed in the top-3 texts sold in the three large English speaking markets. Also listed twice is Jeffery Wooldridge's  Econometric Analysis of Cross Section and Panel Data, which suggests that longitudinal analysis may be in vogue.

Chris Baum’s  An Introduction to Modern Econometrics Using Stata is a bestseller in the UK. This book is a must have for applied econometricians who may not be interested in inverting matrices or computing  gradients and Hessians!

Who digs Econometrics?

Econometrics is a hard science. it involves advance calculus and thus it is not appealing most interested in the American or Indian or some other Idol.

Globally, an interesting landscape emerges for the interest in econometrics. Based on the searches performed on Google between July 2009 and July 2010, Washington, DC, appears to demonstrate the highest interest in econometrics as is evidenced by the normalized searches for the word econometrics performed from Washington, DC.

Singapore comes second in the interest in econometrics. However, being a financial and academic hub, it should not be much of a surprise. New York, with the abundance of financial services firms and higher education institutions comes third. Also, Bill Greene, author of Econometric Analysis, and Robert F. Engle, Noble Laureate in Economics for ARCH and GARCH models, both reside and teach in New York.

Delhi at 4th suggests the fast and sustained rise of the academia in India. New Delhi is the federal capital of India and is also home to a large number of institutions of higher learning. Econometrics must be important to them all.

With David Hensher in Sydney, I am not surprised at Sydney being 5th. There is a growing interest in discrete choice econometrics in Sydney. Toronto on 6th may not be a big cause to celebrate until one realizes that London is trailing Toronto at 7th.

In summary, the top ten cities interested in econometrics are the cities of knowledge and econometrics is the tool of choice for many a researchers.




The following graph is live and presents the phrases that internet surfers have used while searching for econometrics. Notice that Jeffery Wooldridge, author of the famous text Introductory Econometrics, is listed among the top ten searches for econometrics.

Statistics Canada, Dr. Munir Sheikh, and the long form

Yesterday was indeed a sad day for public servants in Canada when one of their finest had to resign in order to best serve his country. Dr. Munir Sheikh, in his resignation letter, has offered Statistics Canada’s assessment that a voluntary survey cannot be a substitute for the mandatory survey.

The controversy started with the federal government trying to change the mandatory nature of the long form of the census to a voluntary one. The Industry Minister implied all along that his decision was backed by the world reputed Statistics Canada. Dr. Sheikh’s resignation letter (which was hurriedly removed from the agency’s website today, but we had preserved it yesterday) however has set the record straight.

Industry Minister Tony Clement claimed multiple times that Statistics Canada was in agreement with his decision to change the survey form. Munir Sheikh’s resignation proves that this was never the case.  We believe that Minister Clement should therefore resign as a minister and as a member of the Parliament for misrepresenting and maligning one of Canada’s finest institution.

It was only a few weeks ago that Minister  Clement asked Canadians to trust Statistics Canada on the controversy surrounding census’s long form. In his resignation letter, the former chief statistician, Dr. Munir Sheikh, has explicitly stated that the voluntary survey cannot be a substitute for the mandatory survey.

I do trust Statistics Canada and after learning the agency’s opinion, I now request Mr. Clement to reverse his decision, which has already been criticised by academics, businesses, and charities.

Mr. Clement had also argued that some of the data collected in the long form was not needed by the government. This is hardly the case unless one assumes that Minister Clement represents all government in Canada. The opposition to his decision by most provincial and all municipal governments suggest that Mr. Clement has exceeded his authority by making a statement on behalf of all tiers of governments and crown corporations, who rely on the census to deliver critical services to Canadians.

Minister Clement has further argued that the decision to change mandatory long form census to voluntary was based on the government’s belief that it should not ask intrusive questions of Canadians. This gives the impression that Prime Minister Stephen Harper’s government respects individual rights and freedoms.

The reality, however, is quite different. This is the same government that refused to spend a dime of foreign aid to provide safe abortions to women in developing countries. It’s a government so blinded by its ideology that it felt entitled to tinker with women’s health. This government is not even remotely concerned about individual’s rights and freedoms.

It is imperative for Canadians to speak out to prevent this government from doing irrevocable damage to Statistics Canada and to ensure that informed decision making practices, which rely on Census data, continue in Canada.

Lastly, Minister Clement is wrong in defending his government’s unconsidered decision to change the long-form census survey from mandatory to voluntary.  The fact that some may find it onerous to respond to long-form census survey is not a sufficient reason to change the nature of Canada’s census.

It also sets a wrong precedent for Statistics Canada who conducts numerous other industry surveys.  The Canadian businesses will not be far behind asking for exemption for the same reason from Statistics Canada’s industry surveys.

The Minister is also misguided in his assertion that the voluntary survey will have a larger unbiased response. The US Census Bureau in 2003 compared the response rates for the mandatory American Community Survey by sending a portion of the respondents the same survey but indicating that their response was voluntary. The change from mandatory to voluntary survey resulted in a huge 20.7% decline in response rate.

Lastly, the US Census Bureau also concluded that the voluntary survey would be much more expensive ($59.2 million more in 2005 dollars) to conduct than the mandatory survey.

Canadians will be well-served if Mr. Clement would come to his senses and leave the census as is. Or better, he should simply resign.

Are housing prices falling in major urban markets in Canada?

A quick answer is yes. However, a detailed look at the graph below, which presents a month-by-month account of seasonally adjusted housing prices in the major urban markets in Canada, suggests that housing prices have faltered as of April 2010 in three major urban markets, namely Toronto, Calgary, and Edmonton, which have also pushed the overall average housing price in Canada lower in the recent months.

The graph presents seasonally adjusted housing prices for the major urban markets in Canada including Vancouver, Calgary, Edmonton, Saskatoon, Toronto, Montreal (truncated), and the average value for Canada.

The most inflated housing market in Canada is that of Vancouver, which is going strong and is only $30,000 shy of reaching the $700,000 average housing price landmark. Since March 2009, the Vancouver housing market has appreciated by $172,000. This is certainly an unsustainable pace of price appreciation. At $11,450 per month in average price appreciation the housing market in Vancouver has certainly outpaced any growth or gain in wages. Thus, a mild decline in housing prices in Vancouver should be forthcoming unless offshore investors yet again bailout the inflated prices in Vancouver.

The average annual price appreciation in Toronto at $6,700 from February 2009 to March 2010 is certainly much less hyped than the one observed in Vancouver. Housing price losses should therefore be milder in Toronto than in Vancouver.

Canada’s housing markets, some reflections

The Canadian housing markets, as of late, have alarmed many a households and economists in Canada. Unlike the United States, where the housing market struggles even today to dig its way out of the housing slump, the Canadian housing markets started to recover as early as in January 2009. The sudden reversal in housing fortunes in Canada is indeed a concern for the market watchers and seller households searching for home buyers.

The Canadian housing recession was relatively short-lived. The nominal housing prices after rising steadily since 1996 dropped only  in 2008. Starting January 2009, the housing prices in Canada started their upward climb, which  continued until April 2010. However, the past few months have revealed a dismal state of affairs in the Canadian housing markets where prices have started to decline again. This is happening during summer months when housing markets in Canada usually heat up.  It is the time when multiple bids become the norm.

As the housing prices decline month after month, economists and government policy makers wonder if this is going to be a double-dip recession.

I have argued previously that the rise in housing prices in Canada in 2009 could be attributed to constrained supply rather than other market fundamentals (such as interest rates, wages, and unemployment), which have not changed much during the recent months.  The support for this argument could be found in the above graph.

In November 2008, there was an abundant supply of resale housing in the market.  The sales to listing ratio stood at 0.38, suggesting that there were 2.7 homes listed for each sale.  However, this also suggests that more sellers were active in the market than buyers.  This resulted in a drop in housing prices, which in turn brought in more buyers to the market. 

As the buyers returned to the market, the sales to listing ratio started climbing upwards from a low of 0.38 in November 2008 to 0.64 in June 2009.  In a short span of seven months, the market turned from abundant supply to constrained supply.  For every home sold in June 2009 there were only 1.5 homes listed in the resale market.  The immediate effect of constrained supply in the housing market was a steady rise in housing prices that began in January 2009 and lasted until April 2010.

As the prices climbed swiftly in 2009, they soon reached the point where buyers withdrew  from the market.  By October 2009, the sales to listing ratio had peaked.  Starting November 2009 the sales to listing ratio declined steadily and did not reverse the trend.  As more sellers entered the market than buyers, it was only a matter of time that housing prices would also reverse their direction and start declining.

Given the fact that housing markets take time to react to new information, it should come as no surprise that the changes in the demand and supply of new housing influence housing prices with a lag of few months.