Apple didn’t make an arbitrary decision to withhold Retina on the Mini to save money, upsell more buyers to the iPad 4, or “force” the first generation of iPad Mini owners to upgrade next year. They chose not to ship a Retina iPad Mini because it would be significantly worse than the previous iPads in very important factors.
I agree with Marco's assessment that Apple did not want to release an iPad Mini that would not meet specific criteria that would be expected of the device, namely being significantly lighter and thinner than the full-sized iPad while retaining comparable battery life (a "full day of use"). The inclusion of a 2048x1536 resolution 7.9-inch screen would have necessitated a larger battery, more RAM (shared between running apps and display buffer), and a more powerful SoC to process all those pixels, adding to the iPad Mini's weight. (The battery wouldn't need to be the same massive 42whr battery in the iPad 3/4 because the LCD would be physically smaller and therefore wouldn't need as much backlighting.) The inclusion of a 2048x1536 (326ppi) screen in this first iPad Mini would have the following consequences for users and Apple:
- Retina LCD panel: adds cost.
- Faster processor: requires more power, adds cost.
- More RAM: requires more (standby) power, adds cost.
- Higher capacity battery: weighs more, takes more physical volume, adds cost.
For Apple, the added monetary cost of all of these components would make the iPad Mini significantly less profitable than it is now--remember that Apple head bean counter Peter Oppenheimer said in Apple's latest financials call that the iPad Mini already has a smaller gross margin than its other iOS products. (currently at the top of its cost curve, factoring in R&D costs in addition to materials and construction). So from the perspective of keeping margins consistent (Apple likes to make money) and not compromising in size/weight/battery life, it makes sense that this year's iPad Mini has a 1024x768 display.
So is a Retina iPad Mini guaranteed next fall?
Hardly. Apple would have to overcome both the technical and financial hurdles that prevented them from releasing a Retina iPad Mini this year. But there are a few signs that point favorably to Apple being able to at least technically make a Retina iPad Mini next year without making it much heavier, heavier, or need to be charged more often than this year's. The secrets lie with the new iPod Touch. This thin and light device--21 playing cards thick, the weight of one full Altoids tin--is remarkable because it has the exact same 1136x640 resolution 4-inch screen as the new iPhone 5. Given that the iPad Mini's screen is basically a larger cut of the non-retina one found in the iPhone 3GS, a larger panel cut from the same 326ppi iPod Touch/iPhone 5 screen is what Apple would need for a Retina iPad Mini. Arrange four iPod Touch's in a 2x2 grid and you can approximate the size and weight of next years iPad Mini (not exactly the same, given the different aspect ratios of 16:9 vs 4:3). And that would still only be 352 grams (four times 88 grams). The iPad Mini is 302 grams now (Wi-Fi model), so that's pretty close. Whether or not Apple can manufacture such a device and keep its profit margins without raising the $330 base price is the larger question. Keep in mind that Apple has never released a newer version of an iOS device and also raised the pricing matrix. Also, Apple really likes making money.
One of Marco's points that I do disagree with:
We can see that a Retina iPad screen is a much bigger power hog than a non-Retina screen of the same size. That’s why the iPad 3 needed to be thicker and heavier than the iPad 2, and why it takes so long to charge: the battery is huge. The iPad 4 has roughly the same size, weight, and battery as the iPad 3, so we know that technological progress hasn’t been able to meaningfully shrink it yet.
I don't think that from a technological perspetive, Apple wasn't able to meaningfully shrink the iPad's size and weight in the iPad 4. The die shrink in the A6X alone would have reduced the power requirements, so Apple could've redesigned the iPad with a smaller battery instead of going for 2x performance (no one I know thinks the iPad 3 was slow by any measure). I think Apple could have at least made the iPad 4 the same size and weight as the iPad 2. A more likely reason for keeping the full size iPad the same size and weight as the 3rd gen version is to accentuate the thinness and lightness of the iPad Mini, while also giving Apple R&D enough time to design a truly significantly thinner and lighter full-size iPad for next fall. It looks like the full-size iPad is on the iPhone upgrade path--they should've just called the iPad 4 the iPad 3S.
John Gruber, in the latest episode of his The Talk Show podcast, makes an even more farfetched hypothesis: Apple kept the size and weight between the iPad 3 and 4 the same to protect the interests of accessory makers, so they didn't have to scramble to redesign their docks and sleeves for a new full-sized iPad for this Holiday season.