If you’re keeping a close eye on the U.S. economy—and who isn’t, these days?—you probably noticed yesterday’s news that retail sales increased in August for the second month in a row. But you may not have noticed that Wolfram|Alpha is now picking up these Department of Commerce reports as soon as they are released, and allows you to explore and compute U.S. retail sales data so you can better understand these trends.

Try simply asking Wolfram|Alpha about “U.S. retail sales” and you’ll see the latest monthly figure, along with automatic computations of that number as a per capita value and as a fraction of total U.S. GDP, as well as the annual growth rate for overall retail sales. To filter out the seasonal variation in many sales categories, you can also ask for “seasonally adjusted retail sales“—which more clearly shows the retail sector’s dramatic plunge in late 2008.

You can also explore trends in individual retail categories (click “More” in the “Retail sales categories” pod for a detailed list), such as clothing stores or electronic shopping and mail order houses.

Or you can mash up this retail sales data with other economic data in Wolfram|Alpha. Try comparing retail sales at building-supply dealers with housing starts, for example, or retail sales at jewelry stores compared with civilian unemployment. (Note that advance figures for August aren’t available for all individual retail categories, so Wolfram|Alpha will default to the latest available values.) More »

Wolfram|Alpha launched with an extensive database of United States economic data, derived from the Federal Reserve Bank’s FRED database. Over the past year, we’ve continued to improve our handling of this data in a variety of ways—teaching Wolfram|Alpha to return more related statistics along with any specific result, improving our linguistic abilities so we can answer more complex questions, and increasing the frequency with which we update this data. Wolfram|Alpha is now refreshing its collection of FRED-derived data on a daily basis, so you can always access the latest available data on the national economy.

We’ve also begun to expand our coverage of economic data for smaller geographic areas in the United States, starting with state-level statistics. This means Wolfram|Alpha users can now query for the latest available information on a variety of economic topics, including gross state product, unemployment, health insurance coverage, and housing-related data.

As always, one of the strengths of Wolfram|Alpha is that it allows you to compare and analyze different pieces of data—and with this data set, you can quickly uncover strong correlations between various economic properties. It’s easy to see that the house price index tends to move together with employment and state tax collections. You can also use Wolfram|Alpha to run simple calculations of productivity in U.S. states, or to find out a given state’s share of the national economy or workforce.

To make it even easier to explore this data, you can also use the new Wolfram|Alpha Widget Builder to create simple tools for analyzing and comparing the economic properties of states. To get you started, here’s a small selection of widgets focused on US state economies—ranging from the serious to the slightly silly. Try them out:

Feel free to customize and share these on Twitter or Facebook, in your blog, or anywhere else—and let us know in the comments if you create any useful new widgets of your own.

This Sunday, July 11, is World Population Day—an event established in 1989 by the United Nations to raise awareness of global population issues. This year, the emphasis is on the 2010 World Population and Housing Census Programme and the importance of collecting, analyzing, and disseminating data in a way that supports good health and social policy development.

In the past few months, we’ve added a variety of international data sets to Wolfram|Alpha, such as data on food consumption and worldwide health indicators. But Wolfram|Alpha launched with an enormous collection of global socioeconomic data, much of it from the UN and other authoritative repositories of international statistics, and we’ve continued to expand and curate that collection.

As we’ve said before, we’re committed to “democratizing data”—to making it easier for everyone to access and understand the wealth of important data produced by a multitude of sources. For good examples of our own ability to analyze and disseminate relevant socioeconomic data, try some of the following queries pertaining to topics from past and present World Population Days:

• All countries’ population in 1900 »
• All countries’ population in 2050 »
• Poverty in all countries »
• Female literacy in all countries »
• Life expectancy in all countries »

We’ll soon be introducing some new functionality that will give “power users” the ability to do more advanced analysis and comparison of properties between groups of countries, and in other knowledge domains. And as always, if you’d like to see additional data in Wolfram|Alpha, please send us your suggestions.

PS: If you’re interested in the absolute latest information on world population, try asking Wolfram|Alpha for the current world population. Reload that page in your browser a few times and see how fast that number is going up!

In any news report about the Deepwater Horizon oil spill, a lot of statistics get thrown around—mainly about the rate at which oil has been spewing out of a pipe on the floor of the Gulf of Mexico. Recent estimates put the flow up to 60,000 barrels per day, but it’s hard for most of us to comprehend exactly what that number means. Wolfram|Alpha has always been able to provide some useful comparisons for any quantity you care to input, and can easily tell you that 60,000 barrels of oil is roughly equivalent to 3.8 times the volume of an Olympic-sized swimming pool.

With the recent addition of data on production and consumption of energy resources in every country, Wolfram|Alpha can also give you a more precise socioeconomic context for numbers like this. Try “60000 barrels per day / US crude oil production”, for example, and you’ll learn that the daily output of the leak is a little less than 1.2% of total crude oil production per day for the United States.

You can also get a better sense of global production or consumption of petroleum products, as well as information on coal and natural gas.

Because each of these energy resources is measured in different units, it can be difficult to understand exactly how they compare to one another—so Wolfram|Alpha can also compute the energy equivalents of each resource, measured in quadrillions of BTUs. For example, you can compare the United States consumption of energy from coal, natural gas, and petroleum on a single scale to better visualize the relative importance of each resource. More »

By the time a new feature or data set is released for public consumption in Wolfram|Alpha, it has already been through a long process of analysis, curation, and design… but even after all of that, we still have our share of “D’oh!” moments at the eleventh hour.

Our latest forehead-smacker was pointed out to us just as we were about to announce the release of historical Academy Awards data. Fortunately, Wolfram|Alpha is flexible enough that we were able to implement a quick, partial fix before this year’s Oscars ceremony—but we also had to go back and do some more substantial work so this data is presented with absolute clarity.

So what was the problem? We had taken for granted the idea that when users typed in “2010 Academy Awards”, they’d expect to see people who won Oscars at this year’s ceremony… and then we just counted backward from there, to the first Oscars ceremony in 1929. But as it was pointed out, if you ask “who won the Oscar for best supporting actor in 2005”, you might want to know about films released in 2005, not films honored at the 2005 Academy Awards ceremony. So now when you ask for information about Oscars we assume you mean the year of the award ceremony, but for most years you can also click on a single link in the assumption pod to interpret your input as referring to year of film release instead.

We’ve also cleaned up the presentation of some quirks in Oscar history, including the unique case of the Academy Awards in 1930—when there were actually two ceremonies in a single year, one for films released in 1928–29, and one for films released in 1929–30:

For other early Academy Awards, we still assume that input refers to the year of the ceremony, but we’ve added a footnote that provides more details about releases date for the films honored in that year. And we’ve added a few other little features, like the ability to handle queries like “best actor at the 42nd Academy Awards”, or to ask about specific dates of Academy Awards ceremonies. More »

Wolfram|Alpha couldn’t do your taxes for you this year, but we did just wrap up a quick project to add some interesting historical tax statistics. Now that all of our U.S. users have filed their taxes (we hope), they can explore IRS data about individual income taxes, broken down by adjusted gross income (AGI), from 1996 to 2007—the latest year for which the IRS has released statistics broken down by AGI. Users can also investigate less-detailed data about sources of individual taxable income from 1916 to 2007.

The basic input for this new dataset is simply an income, such as “AGI $35000”—type it in, and Wolfram|Alpha matches that input to a specific AGI bracket (in this case, $30,000–$40,000) and calculates a broad range of statistics.

First, the average effective federal tax rate, which is calculated by dividing total tax receipts in this bracket by total adjusted gross income:

Next, the average tax paid in the input’s bracket—which in this case dropped by nearly 50% over the decade covered by this dataset. You’ll also note that nearly a quarter of all tax returns in this AGI bracket had no tax due:

Third, average exemptions and deductions for all taxpayers in the input’s bracket. In this case, those increased by nearly $4,000 over this period, and in 2007 accounted for an average of 48% of AGI:

For some particularly interesting numbers, try asking about high income brackets (average tax on AGI $400k, average exemptions and deductions on $5 million) and very low brackets (average tax on AGI $500). More »

Since Wolfram|Alpha‘s launch in May 2009, one of its most talked-about features has been its ability to compute specific answers to questions about math, chemistry, economics, demographics, and much more. But as its knowledge base continues to grow, it’s also able to highlight interesting and useful connections between data sets, and to reveal information that you might not think to ask for on your own.

One of the coolest examples of this is our recently enhanced relocation calculator. For several months, we’ve been able to answer simple questions about the relative cost of living in various United States cities and metropolitan areas. If you told Wolfram|Alpha that you were relocating from Seattle to Miami with a salary of $35000, you’d get a comparison of the relative cost of groceries, housing, and other expenses in each city, plus an estimate of the salary required to maintain a comparable standard of living in your destination city. On its own, this is a useful little calculator—but it’s also something that dozens of other websites could do.

But because Wolfram|Alpha knows tons of other details about any given city, our relocation calculator can now do things that no other site can. In addition to salary and cost-of-living comparisons, you now get comparisons of each city’s population, median home sale prices, unemployment rates, crime rates, sales taxes, traffic congestion, and climate—a useful sampling of current and historical comparative data for anyone contemplating a move.

We’ll highlight similar enhancements as they are released. And as always, we welcome your suggestions for new data, or new ways of looking at existing data, in any domain covered by Wolfram|Alpha.

With the 2010 Academy Awards coming up this Sunday, we’re happy to announce that Wolfram|Alpha is now able to answer questions about every Oscar nomination and award since the first ceremony in 1929. You might be surprised by some of the things you see in the earliest lists: yes, acting awards were bestowed for multiple performances in a given year; the Academy made a distinction between movies that were merely “unique and artistic” and those that were truly “outstanding”; and like the current Golden Globes (we’ll tackle them soon), separate awards were given for dramatic and comedic films.

You can dive into this data in practically any way you want. Curious about a particular film? Try “Academy Award nominations for Forrest Gump“. Or maybe you’re curious about the past performance of a perennial front-row Oscar celebrity?

Ask about a specific award, like “best actor oscars“, and you’ll get a historical list of all winners for that category. But ask about “best actor in 2004“, and Wolfram|Alpha will serve up a detailed cross-section of data relevant to that award—the winner, other nominees, and other Oscar nominations and awards for both the winner and the film he appeared in. More »

Happy birthday, George Washington! In case you’d forgotten, President’s Day in the United States isn’t actually celebrated on George Washington’s birthday: since 1971, it has fallen on the 3rd Monday in February, which means it’s always at least one day short of the first president’s actual birthday, February 22.

As you might imagine for a man referred to as “the father of the country,” the name “Washington” has taken on a life of its own—and as such, it provides a good opportunity to see how Wolfram|Alpha deals with cases where a single word can be interpreted in many different ways.

Type “Washington” on its own, and you’ll learn that the word could refer to a city, a U.S. state, a surname, a specific person, or a given name. For users in the United States, Wolfram|Alpha will assume you’re talking about the nation’s capital, and then give a list of alternate cities ranked by a combination of population, distance from your current location, and general popularity. But if you’re in the United Kingdom, the default assumption will be a place closer to home:

When you ask more-specific questions about “Washington”, Wolfram|Alpha is usually able to make even-more-intelligent assumptions about which Washington you really want know about. Ask for “distance from seattle to washington” and you’ll get the great-circle distance between two cities. Try to “compare virginia and washington“, and you’ll get a stat-by-stat comparison of the two U.S. states. Ask Wolfram|Alpha “when was Washington born?” and the result is the first U.S. president’s birthday; try “washingtons in 1900” and you’ll discover that about 28 U.S. residents were given that first name that year, or ask about “washington as a last name” and Wolfram|Alpha will reveal that more than 160,000 people had that last name in the 2000 U.S. Census. More »

If all you know about Groundhog Day is what you learned from watching the Bill Murray movie, well… you’re actually quite well informed. The good people of Punxsutawney, Pennsylvania really do gather at Gobbler’s Knob each February 2 to find out whether a 20-pound groundhog named Punxsutawney Phil will see his shadow, thus foretelling six more weeks of winter—or not, foretelling an early spring.

In more than 120 years of predictions, there have only been 15 occasions on which Phil hasn’t seen his shadow. The National Climatic Data Center has estimated Phil’s accuracy rate at around 39%, but true Phil fans (or skeptics) can do their own analysis of Phil’s track record with Wolfram|Alpha.

Let’s take 1950, for example: according to Punxsutawney’s “Inner Circle,” Phil did not see his shadow that year. Ask about “punxsutawney, pennsylvannia weather feb. 2 1950,” and you discover that practically the entire day was overcast and foggy: not good conditions for a giant rodent to see his shadow. But an early spring? Check the results for “punxsutawney, pennsylvannia weather february 1950” and it’s hard to overlook the plunging temperatures and snowfall in the latter part of the month. Sorry, Phil. More »